Courses with examples to learn how to work with Excel files in Python
Working with Excel files in Python
The Tutorial Materials
These can be obtained by CD, USB drive or downloaded from here:
The Website
The best place to start when working with Excel files in python is the website:
Introduction
This tutorial covers the following libraries: xlrd
• reading data and formatting from .xls files
• this tutorial covers version 0.7.1 • API documentation can be found at:
◦xlwt
• writing data and formatting to .xls files
• this tutorial covers version 0.7.2
• Incomplete API documentation can be found at:
• Fairly complete examples can be found at ◦
• a collection of utilities using both xlrd and xlwt:
◦ copying data from a source to a target spreadsheet ◦ filtering data from a source to a target spreadsheet
• this tutorial covers version 1.3.0 and above.
• Documentation and examples can be found at:
There are still reasons why automating an Excel instance via COM is necessary:
• manipulation of graphs
• rich text cells
• reading formulae in cells
• working with macros and names
• the more esoteric things found in .xls files
Installation
There are several methods of installation available. While the following examples are for xlrd, the exact same steps can be used for any of the three libraries.
Install from Source On Linux:
$ tar xzf
$ cd xlrd-0.7.1
$ python install
NB: Make sure you use the python you intend to use for your project.
On Windows, having used WinZip or similar to unpack :
C:\> cd xlrd-0.7.1
C:\xlrd-0.7.1> \Python26\python install
NB: Make sure you use the python you intend to use for your project.
Install using Windows Installer
On Windows, you can download and run the installer.
Beware that this will only install to Python installations that are in the windows registry.
Install using EasyInstall
This cross-platform method requires that you already have EasyInstall installed. For more information on this, please see:
easy_install xlrd
Installation using Buildout
Buildout provides a cross-platform method of meeting the python package dependencies of a project without interfering with the system python.
Having created a directory called mybuildout, download the following file into it:
• *checkout*
Now, create a file in mybuildout called containing the following:
[buildout] parts = py versions = versions [versions] xlrd=0.7.1 xlwt=0.7.2 xlutils=1.3.2 [py] recipe = eggs = xlrd xlwt xlutils interpreter = py |
NB: The versions section is optional Finally, run the following:
$ python
$ bin/buildout
These lines:
• initialise the buildout environment
• run the buildout. This should be done each time dependencies change. Now you can do the following:
$ bin/py
Buildout lives at
Reading Excel Files
All the examples shown below can be found in the xlrd directory of the course material.
Opening Workbooks
Workbooks can be loaded either from a file, an object or from a string:
from mmap import mmap,ACCESS_READ from xlrd import open_workbook print open_workbook('') with open('', 'rb') as f: print open_workbook( file_contents=mmap(f.fileno(),0,access=ACCESS_READ) ) aString = open('','rb').read() print open_workbook(file_contents=aString) |
Navigating a Workbook
Here is a simple example of workbook navigation:
from xlrd import open_workbook wb = open_workbook('') for s in wb.sheets(): print 'Sheet:',s.name for row in range(s.nrows): values = [] for col in range(s.ncols): values.append(s.cell(row,col).value) print ','.join(values) print |
The next few sections will cover the navigation of workbooks in more detail.
Introspecting a Book
The object returned by open_workbook contains all information to do with the workbook and can be used to retrieve individual sheets within the workbook.
The nsheets attribute is an integer containing the number of sheets in the workbook. This attribute, in combination with the sheet_by_index method is the most common way of retrieving individual sheets.
The sheet_names method returns a list of unicodes containing the names of all sheets in the workbook. Individual sheets can be retrieved using these names by way of the sheet_by_name function.
The results of the sheets method can be iterated over to retrieve each of the sheets in the workbook.
The following example demonstrates these methods and attributes:
from xlrd import open_workbook book = open_workbook('') print book.nsheets for sheet_index in range(book.nsheets): print book.sheet_by_index(sheet_index) print book.sheet_names() for sheet_name in book.sheet_names(): print book.sheet_by_name(sheet_name) for sheet in book.sheets(): print sheet |
objects have other attributes relating to the content of the workbook that are only rarely useful:
• codepage
• countries
• user_name
If you think you may need to use these attributes, please see the xlrd documentation.
Introspecting a Sheet
The xlrd.sheet.Sheet objects returned by any of the methods described above contain all the information to do with a worksheet and its contents.
The name attribute is a unicode representing the name of the worksheet.
The nrows and ncols attributes contain the number of rows and the number of columns, respectively, in the worksheet.
The following example shows how these can be used to iterate over and display the contents of one worksheet:
from xlrd import open_workbook,cellname book = open_workbook('') sheet = book.sheet_by_index(0) print print sheet.nrows print sheet.ncols for row_index in range(sheet.nrows): for col_index in range(sheet.ncols): print cellname(row_index,col_index),'-', print (row_index,col_index).value |
xlrd.sheet.Sheet objects have other attributes relating to the content of the worksheet that are only rarely useful:
• col_label_ranges
• row_label_ranges
• visibility
If you think you may need to use these attributes, please see the xlrd documentation.
Getting a particular Cell
As already seen in previous examples, the cell method of a Sheet object can be used to return the contents of a particular cell.
The cell method returns an object. These objects have very few attributes, of which value contains the actual value of the cell and ctype contains the type of the cell.
In addition, Sheet objects have two methods for returning these two types of data. The cell_value method returns the value for a particular cell, while the cell_type method returns the type of a particular cell. These methods can be quicker to execute than retrieving the Cell object.
Cell types are covered in more detail later. The following example shows the methods, attributes and classes in action:
from xlrd import open_workbook,XL_CELL_TEXT book = open_workbook('') sheet = book.sheet_by_index(1) cell = (0,0) print cell print cell.value print cell.ctype==XL_CELL_TEXT for i in range(sheet.ncols): print sheet.cell_type(1,i),sheet.cell_value(1,i) |
Iterating over the contents of a Sheet
We've already seen how to iterate over the contents of a worksheet and retrieve the resulting individual cells. However, there are methods to retrieve groups of cells more easily. There are a symmetrical set of methods that retrieve groups of cell information either by row or by column.
The row method and col method return all the Cell objects for a whole row or column respectively.
The row_slice and col_slice methods return a list of Cell objects in a row or column, respectively, bounded by and start index and an optional end index.
The row_types and col_types methods return a list of integers representing the cell types in a row or column, respectively, bounded by and start index and an optional end index.
The row_values and col_values methods return a list of objects representing the cell values in a row or column, respectively, bounded by and start index and an optional end index.
The following examples demonstrates all of the sheet iteration methods:
from xlrd import open_workbook book = open_workbook('') sheet0 = book.sheet_by_index(0) sheet1 = book.sheet_by_index(1) print (0) print (0) print sheet0.row_slice(0,1) print sheet0.row_slice(0,1,2) print sheet0.row_values(0,1) print sheet0.row_values(0,1,2) print sheet0.row_types(0,1) print sheet0.row_types(0,1,2) print sheet1.col_slice(0,1) print sheet0.col_slice(0,1,2) print sheet1.col_values(0,1) print sheet0.col_values(0,1,2) print sheet1.col_types(0,1) print sheet0.col_types(0,1,2) |
Utility Functions
When navigating around a workbook, it's often useful to be able to convert between row and column indexes and the Excel cell references that users may be used to seeing. The following functions are provided to help with this:
The cellname function turns a row and column index into a relative Excel cell reference.
The cellnameabs function turns a row and column index into an absolute Excel cell reference.
The colname function turns a column index into an Excel column name.
These three functions are demonstrated in the following example:
from xlrd import cellname, cellnameabs, colname
print cellname(0,0),cellname(10,10),cellname(100,100) print cellnameabs(3,1),cellnameabs(41,59),cellnameabs(265,358) print colname(0),colname(10),colname(100)
Unicode
All text attributes and values produced by xlrd will be either unicode objects or, in rare cases, ascii encoded strings.
Each piece of text in an Excel file written by Microsoft Excel is encoded into one of the following:
• Latin1, if it fits
• UTF_16_LE, if it doesn't find into Latin1
• In older files, by an encoding specified by an MS codepage. These are mapped to Python encodings by xlrd and still results in unicode objects.
In rare cases, other software has been know to write no codepage or the wrong codepage into Excel files. In this case, the correct encoding may need to be specified to open_workbook:
from xlrd import open_workbook book = open_workbook('',encoding='cp1252')
Types of Cell
We have already seen the cell type expressed as an integer. This integer corresponds to a set of constants in xlrd that identify the type of the cell. The full set of possible cell types is listed in the following sections.
Text
These are represented by the xlrd.XL_CELL_TEXT constant.
Cells of this type will have values that are unicode objects.
Number
These are represented by the xlrd.XL_CELL_NUMBER constant. Cells of this type will have values that are float objects.
Date
These are represented by the xlrd.XL_CELL_DATE constant.
NB: Dates don't really exist in Excel files, they are merely Numbers with a particular number formatting.
xlrd will return xlrd.XL_CELL_DATE as the cell type if the number format string looks like a date.
The xldate_as_tuple method is provided for turning the float in a Date cell into a tuple suitable for instantiating various date/time objects. This example shows how to use it:
from datetime import date,datetime,time from xlrd import open_workbook,xldate_as_tuple book = open_workbook('') sheet = book.sheet_by_index(0) date_value = xldate_as_tuple((3,2).value,book.datemode) print datetime(*date_value),date(*date_value[:3]) datetime_value = xldate_as_tuple((3,3).value,book.datemode) print datetime(*datetime_value) time_value = xldate_as_tuple((3,4).value,book.datemode) print time(*time_value[3:]) print datetime(*time_value) |
Caveats:
• Excel files have two possible date modes, one for files originally created on Windows and one for files originally created on an Apple machine. This is expressed as the datemode attribute of objects and must be passed to xldate_as_tuple.
• The Excel file format has various problems with dates before 3 Jan 1904 that can cause date ambiguities that can result in xldate_as_tuple raising an XLDateError.
• The Excel formula function DATE()can return unexpected dates in certain circumstances.
Boolean
These are represented by the xlrd.XL_CELL_BOOLEAN constant. Cells of this type will have values that are bool objects.
Error
These are represented by the xlrd.XL_CELL_ERROR constant.
Cells of this type will have values that are integers representing specific error codes.
The error_text_from_code function can be used to turn error codes into error messages:
from xlrd import open_workbook,error_text_from_code book = open_workbook('') sheet = book.sheet_by_index(0) print error_text_from_code[(5,2).value] print error_text_from_code[(5,3).value] |
For a simpler way of sensibly displaying all cell types, see xlutils.display.
Empty / Blank
Excel files only store cells that either have information in them or have formatting applied to them. However, xlrd presents sheets as rectangular grids of cells.
Cells where no information is present in the Excel file are represented by the xlrd.XL_CELL_EMPTY constant. In addition, there is only one “empty cell”, whose value is an empty string, used by xlrd, so empty cells may be checked using a Python identity check.
Cells where only formatting information is present in the Excel file are represented by the xlrd.XL_CELL_BLANK constant and their value will always be an empty string.
from xlrd import open_workbook,empty_cell print empty_cell.value book = open_workbook('') sheet = book.sheet_by_index(0) empty = (6,2) blank = (7,2) print empty is blank, empty is empty_cell, blank is empty_cell book = open_workbook('',formatting_info=True) sheet = book.sheet_by_index(0) empty = (6,2) blank = (7,2) print empty.ctype,repr(empty.value) print blank.ctype,repr(blank.value) |
The following example brings all of the above cell types together and shows examples of their use:
from xlrd import open_workbook def cell_contents(sheet,row_x): result = [] for col_x in range(2,sheet.ncols): cell = (row_x,col_x) result.append((cell.ctype,cell,cell.value)) return result sheet = open_workbook('').sheet_by_index(0) print 'XL_CELL_TEXT',cell_contents(sheet,1) print 'XL_CELL_NUMBER',cell_contents(sheet,2) print 'XL_CELL_DATE',cell_contents(sheet,3) print 'XL_CELL_BOOLEAN',cell_contents(sheet,4) print 'XL_CELL_ERROR',cell_contents(sheet,5) print 'XL_CELL_BLANK',cell_contents(sheet,6) print 'XL_CELL_EMPTY',cell_contents(sheet,7) sheet = open_workbook( '',formatting_info=True ).sheet_by_index(0) print 'XL_CELL_TEXT',cell_contents(sheet,1) print 'XL_CELL_NUMBER',cell_contents(sheet,2) print 'XL_CELL_DATE',cell_contents(sheet,3) print 'XL_CELL_BOOLEAN',cell_contents(sheet,4) print 'XL_CELL_ERROR',cell_contents(sheet,5) print 'XL_CELL_BLANK',cell_contents(sheet,6) print 'XL_CELL_EMPTY',cell_contents(sheet,7) |
Names
These are an infrequently used but powerful way of abstracting commonly used information found within Excel files.
They have many uses, and xlrd can extract information from many of them. A notable exception are names that refer to sheet and VBA macros, which are extracted but should be ignored.
Names are created in Excel by navigating to Insert > Name > Define. If you plan to use xlrd to extract information from Names, familiarity with the definition and use of names in your chosen spreadsheet application is a good idea.
Types
A Name can refer to:
• A constant
◦ CurrentInterestRate = 0.015
◦ NameOfPHB = “Attila T. Hun”
• An absolute (i.e. not relative) cell reference
◦ CurrentInterestRate = Sheet1!$B$4
• Absolute reference to a 1D, 2D, or 3D block of cells
◦ MonthlySalesByRegion = Sheet2:Sheet5!$A$2:$M$100
• A list of absolute references
◦ Print_Titles = [row_header_ref, col_header_ref])
Constants can be extracted.
The coordinates of an absolute reference can be extracted so that you can then extract the corresponding data from the relevant sheet(s).
A relative reference is useful only if you have external knowledge of what cells can be used as the origin. Many formulas found in Excel files include function calls and multiple references and are not useful, and can be too hard to evaluate. A full calculation engine is not included in xlrd.
Scope
The scope of a Name can be global, or it may be specific to a particular sheet. A Name's identifier may be re-used in different scopes. When there are multiple Names with the same identifier, the most appropriate one is used based on scope. A good example of this is the built-in name Print_Area; each worksheet may have one of these.
Examples:
name=rate, scope=Sheet1, formula=0.015 name=rate, scope=Sheet2, formula=0.023 name=rate, scope=global, formula=0.040
A cell formula (1+rate)^20 is equivalent to 1.015^20 if it appears in Sheet1 but equivalent to 1.023^20 if it appears in Sheet2, and 1.040^20 if it appears in any other sheet.
Usage
Common reasons for using names include:
• Assigning textual names to values that may occur in many places within a workbook ◦ eg: RATE = 0.015
• Assigning textual names to complex formulae that may be easily mis-copied
◦ eg: SALES_RESULTS = $A$10:$M$999
Here's an example real-world use case: reporting to head office. A company's head office makes up a template workbook. Each department gets a copy to fill in. The various ranges of data to be provided all have defined names. When the files come back, a script is used to validate that the department hasn't trashed the workbook and the names are used to extract the data for further processing. Using names decouples any artistic repositioning of the ranges, by either head office template-designing user or by departmental users who are filling in the template, from the script which only has to know what the names of the ranges are.
In the examples directory of the xlrd distribution you will find which has examples of most of the non-macro varieties of defined names. There is also which shows how to use the name lookup dictionaries, and how to extract constants and references and the data that references point to.
Formatting
We've already seen that open_workbook has a parameter to load formatting information from Excel files. When this is done, all the formatting information is available, but the details of how it is presented are beyond the scope of this tutorial.
If you wish to copy existing formatted data to a new Excel file, see and xlutils.filter.
If you do wish to inspect formatting information, you'll need to consult the following
colour_map font_list format_list | palette_record style_name_map xf_list |
attributes of the following classes:
format_map
xlrd.sheet.Sheet
cell_xf_index rowinfo_map colinfo_map computed_column_width default_additional_space_above default_additional_space_below | default_row_height_mismatch default_row_hidden defcolwidth gcw merged_cells standard_width |
default_row_height xf_index
Other Classes
In addition, the following classes are solely used to represent formatting information:
xlrd.sheet.Rowinfo xlrd.sheet.Colinfo xlrd.formatting.Format | xlrd.formatting.XFAlignment xlrd.formatting.XFBackground xlrd.formatting.XFBorder xlrd.formatting.XFProtection |
Working with large Excel files
If you're working with particularly large Excel files then there are two features of xlrd that you should be aware of:
• The on_demand parameter can be passed as True to open_workbook resulting in worksheets only being loaded into memory when they are requested.
• objects have an unload_sheet method that will unload worksheet, specified by either sheet index or sheet name, from memory.
The following example shows how a large workbook could be iterated over when only sheets matching a certain pattern need to be inspected, and where only one of those sheets ends up in memory at any one time:
from xlrd import open_workbook book = open_workbook('',on_demand=True) for name in book.sheet_names(): if name.endswith('2'): sheet = book.sheet_by_name(name) print sheet.cell_value(0,0) book.unload_sheet(name) |
Introspecting Excel files with
The xlrd source distribution includes a script that is extremely useful for introspecting Excel files without writing a single line of Python.
You are encouraged to run a variety of the commands it provides over the Excel files provided in the course materials.
The following gives an overview of what's available from runxlrd, and can be obtained using python –-help:
[options] command [input-file-patterns] Commands:
2rows Print the contents of first and last row in each sheet
3rows Print the contents of first, second and last row in each sheet
bench Same as "show", but doesn't print -- for profiling biff_count[1] Print a count of each type of BIFF record in the file biff_dump[1] Print a dump (char and hex) of the BIFF records in the file fonts hdr + print a dump of all font objects hdr Mini-overview of file (no per-sheet information)
hotshot Do a hotshot profile run e.g. -f1 hotshot bench bigfile*.xls labels Dump of sheet.col_label_ranges and row for each sheet name_dump Dump of each object in book.name_obj_list names Print brief information for each NAME record ov Overview of file
profile Like "hotshot", but uses cProfile show Print the contents of all rows in each sheet version[0] Print versions of xlrd and Python and exit
xfc Print "XF counts" and cell-type counts -- see code for details
[0] means no file arg
[1] means only one file arg i.e. no pattern
Options:
-h, --help show this help message and exit
-l LOGFILENAME, --logfilename=LOGFILENAME contains error messages
-v VERBOSITY, --verbosity=VERBOSITY
level of information and diagnostics provided -p PICKLEABLE, --pickleable=PICKLEABLE
1: ensure Book object is pickleable (default); 0: don't bother
-m MMAP, --mmap=MMAP 1: use mmap; 0: don't use mmap; -1: accept heuristic
-e ENCODING, --encoding=ENCODING encoding override -f FORMATTING, --formatting=FORMATTING
0 (default): no fmt info 1: fmt info (all cells)
-g GC, --gc=GC 0: auto gc enabled; 1: auto gc disabled, manual
collect after each file; 2: no gc
-s ONESHEET, --onesheet=ONESHEET
restrict output to this sheet (name or index)
-u, --unnumbered omit line numbers or offsets in biff_dump
Writing Excel Files
All the examples shown below can be found in the xlwt directory of the course material.
Creating elements within a Workbook
Workbooks are created with xlwt by instantiating an xlwt.Workbook object, manipulating it and then calling its save method.
The save method may be passed either a string containing the path to write to or a filelike object, opened for writing in binary mode, to which the binary Excel file data will be written.
The following objects can be created within a workbook:
Worksheets
Worksheets are created with the add_sheet method of the Workbook class.
To retrieve an existing sheet from a Workbook, use its get_sheet method. This method is particularly useful when the Workbook has been instantiated by .
Rows
Rows are created using the row method of the Worksheet class and contain all of the cells for a given row.
The row method is also used to retrieve existing rows from a Worksheet.
If a large number of rows have been written to a Worksheet and memory usage is becoming a problem, the flush_row_data method may be called on the Worksheet. Once called, any rows flushed cannot be accessed or modified.
It is recommended that flush_row_data is called for every 1000 or so rows of a normal size that are written to an xlwt.Workbook. If the rows are huge, that number should be reduced.
Columns
Columns are created using the col method of the Worksheet class and contain display formatting information for a given column.
The col method is also used to retrieve existing columns from a Worksheet. Cells
Cells can be written using either the write method of either the Worksheet or Row class.
A more detailed discussion of different ways of writing cells and the different types of cell that may be written is covered later.
A Simple Example
The following example shows how all of the above methods can be used to build and save a simple workbook:
from tempfile import TemporaryFile from xlwt import Workbook book = Workbook() sheet1 = book.add_sheet('Sheet 1') book.add_sheet('Sheet 2') sheet1.write(0,0,'A1') sheet1.write(0,1,'B1') row1 = (1) row1.write(0,'A2') row1.write(1,'B2') (0).width = 10000 sheet2 = book.get_sheet(1) (0).write(0,'Sheet 2 A1') (0).write(1,'Sheet 2 B1') sheet2.flush_row_data() sheet2.write(1,0,'Sheet 2 A3') (0).width = 5000 (0).hidden = True ('') (TemporaryFile()) |
Unicode
The best policy is to pass unicode objects to all xlwt-related method calls.
If you absolutely have to use encoded strings then make sure that the encoding used is consistent across all calls to any xlwt-related methods.
If encoded strings are used and the encoding is not 'ascii', then any Workbook objects must be created with the appropriate encoding specified:
from xlwt import Workbook book = Workbook(encoding='utf-8')
Writing to Cells
A number of different ways of writing a cell are provided by xlwt along with different strategies for handling multiple writes to the same cell.
Different ways of writing cells
There are generally three ways to write to a particular cell:
• Worksheet.write(row_index,column_index,value)
◦ This is just syntactic sugar for (row_index).write(column_index,value)
◦ It can be useful when you only want to write one cell to a row
• Row.write(column_index,value)
◦ This will write the correct type of cell based on the value passed
◦ Because it figures out what type of cell to write, this method may be slower for writing large workbooks
• Specialist write methods on the Row class
◦ Each type of cell has a specialist setter method as covered in the “Types of Cell” section below.
◦ These require you to pass the correct type of Python object but can be faster.
In general, use Worksheet.write for convenience and the specialist write methods if you require speed for a large volume of data.
Overwriting Cells
The Excel file format does nothing to prevent multiple records for a particular cell occurring but, if this happens, the results will vary depending on what application is used to open the file. Excel will display a “File error: data may have been lost” while will show the last record for the cell that occurs in the file. To help prevent this, xlwt provides two modes of operation:
• Writing to the same cell more than once will result in an exception This is the default mode.
• Writing to the same cell more than once will replace the record for that cell, and only one record will be written when the Workbook is saved.
The following example demonstrates these two options:
from xlwt import Workbook book = Workbook() sheet1 = book.add_sheet('Sheet 1',cell_overwrite_ok=True) sheet1.write(0,0,'original') sheet = book.get_sheet(0) sheet.write(0,0,'new') sheet2 = book.add_sheet('Sheet 2') sheet2.write(0,0,'original') sheet2.write(0,0,'new') |
The most common case for needing to overwrite cells is when an existing Excel file has been loaded into a Workbook instance using .
Types of Cell
All types of cell supported by the Excel file format can be written:
Text
When passed a unicode or string, the write methods will write a Text cell.
The set_cell_text method of the Row class can also be used to write Text cells.
When passed a string, these methods will first decode the string using the Workbook's encoding.
Number
When passed a float, int, long, or decimal.Decimal, the write methods will write a Number cell.
The set_cell_number method of the Row class can also be used to write Number cells.
Date
When passed a datetime.datetime, or , the write methods will write a Date cell.
The set_cell_date method of the Row class can also be used to write Date cells.
Note: As mentioned earlier, a date is not really a separate type in Excel; if you don't apply a date format, it will be treated as a number.
Boolean
When passed a bool, the write methods will write a Boolean cell.
The set_cell_boolean method of the Row class can also be used to write Text cells.
Error
You shouldn't ever want to write Error cells!
However, if you absolutely must, the set_cell_error method of the Row class can be used to do so. For convenience, it can be called with either hexadecimal error codes, expressed as integers, or the error text that Excel would display.
Blank
It is not normally necessary to write blank cells. The one exception to this is if you wish to apply formatting to a cell that contains nothing.
To do this, either call the write methods with an empty string or None, or use the set_cell_blank method of the Row class.
If you need to do this for more than one cell in a row, using the set_cell_mulblanks method will result in a smaller Excel file when the Workbook is saved.
The following example brings all of the above cell types together and shows examples use both the generic write method and the specialist methods:
from datetime import date,time,datetime from decimal import Decimal from xlwt import Workbook,Style wb = Workbook() ws = wb.add_sheet('Type examples') (0).write(0,u'\xa3') (0).write(1,'Text') (1).write(0,3.1415) (1).write(1,15) (1).write(2,265L) (1).write(3,Decimal('3.65')) (2).set_cell_number(0,3.1415) (2).set_cell_number(1,15) (2).set_cell_number(2,265L) (2).set_cell_number(3,Decimal('3.65')) (3).write(0,date(2009,3,18)) (3).write(1,datetime(2009,3,18,17,0,1)) (3).write(2,time(17,1)) (4).set_cell_date(0,date(2009,3,18)) (4).set_cell_date(1,datetime(2009,3,18,17,0,1)) (4).set_cell_date(2,time(17,1)) (5).write(0,False) (5).write(1,True) (6).set_cell_boolean(0,False) (6).set_cell_boolean(1,True) (7).set_cell_error(0,0x17) (7).set_cell_error(1,'#NULL!') (8).write( 0,'',Style.easyxf('pattern: pattern solid, fore_colour green;')) (8).write( 1,None,Style.easyxf('pattern: pattern solid, fore_colour blue;')) (9).set_cell_blank( 0,Style.easyxf('pattern: pattern solid, fore_colour yellow;')) (10).set_cell_mulblanks( 5,10,Style.easyxf('pattern: pattern solid, fore_colour red;') ) ('') |
Styles
Most elements of an Excel file can be formatted. For many elements including cells, rows and columns, this is done by assigning a style, known as an XF record, to that element. This is done by passing an xlwt.XFStyle instance to the optional last argument to the various write methods and specialist set_cell_ methods. and xlwt.Column instances have set_style methods to which an xlwt.XFStyle instance can be passed.
XFStyle
In xlwt, the XF record is represented by the XFStyle class and its related attribute classes. The following example shows how to create a red Date cell with Arial text and a black border:
from datetime import date from xlwt import Workbook, XFStyle, Borders, Pattern, Font fnt = Font() = 'Arial' borders = Borders() = Borders.THICK borders.right = Borders.THICK = Borders.THICK borders.bottom = Borders.THICK pattern = Pattern() pattern.pattern = Pattern.SOLID_PATTERN pattern.pattern_fore_colour = 0x0A style = XFStyle() style.num_format_str='YYYY-MM-DD' = fnt style.borders = borders style.pattern = pattern book = Workbook() sheet = book.add_sheet('A Date') sheet.write(1,1,date(2009,3,18),style) ('') |
This can be quite cumbersome!
easyxf
Thankfully, xlwt provides the easyxf helper to create XFStyle instances from human readable text and an optional string containing a number format. Here is the above example, this time created with easyxf:
from datetime import date from xlwt import Workbook, easyxf book = Workbook() sheet = book.add_sheet('A Date') sheet.write(1,1,date(2009,3,18),easyxf( 'font: name Arial;' 'borders: left thick, right thick, top thick, bottom thick;' 'pattern: pattern solid, fore_colour red;', num_format_str='YYYY-MM-DD' )) ('') |
The human readable text breaks roughly as follows, in pseudo-regular expression syntax:
(:( ,)+;)+ This means:
• The text contains a semi-colon delimited list of element definitions.
• Each element contains a comma-delimited list of attribute and value pairs. The following sections describe each of the types of element by providing a table of their attributes and possible values for those attributes. For explanations of how to express boolean values and colours, please see the “Types of attribute” section.
font
bold | A boolean value. The default is False. |
charset | The character set to use for this font, which can be one of the following: ansi_latin, sys_default, symbol, apple_roman, ansi_jap_shift_jis, ansi_kor_hangul, ansi_kor_johab, ansi_chinese_gbk, ansi_chinese_big5, ansi_greek, ansi_turkish, ansi_vietnamese, ansi_hebrew, ansi_arabic, ansi_baltic, ansi_cyrillic, ansi_thai, ansi_latin_ii, oem_latin_i The default is sys_default. |
colour | A colour specifying the colour for the text. The default is the automatic colour. |
escapement | This can be one of none, superscript or subscript. The default is none. |
family | This should be a string containing the name of the font family to use. You probably want to use name instead of this attribute and leave this to its default value. The default is None. |
height | The height of the font as expressed by multiplying the point size by 20. The default is 200, which equates to 10pt. |
italic | A boolean value. The default is False. |
name | This should be a string containing the name of the font family to use. The default is Arial. |
outline | A boolean value. The default is False. |
shadow | A boolean value. The default is False. |
struck_out | A boolean value. The default is False. |
underline | A boolean value or one of none, single, single_acc, double or double_acc. The default is none. |
color_index | A synonym for colour |
colour_index | A synonym for colour |
color | A synonym for colour |
alignment
direction | One of general, lr, or rl. The default is general. |
horizontal | One of the following: general, left, center|centre, right, filled, justified, center|centre_across_selection, distributed The default is general. |
indent | A indentation amount between 0 and 15. The default is 0. |
rotation | An integer rotation in degrees between -90 and +90 or one of stacked or none. The default is none. |
shrink_to_fit | A boolean value. The default is False. |
vertical | One of the following: top, center|centre, bottom, justified, distributed The default is bottom. |
wrap | A boolean value. The default is False. |
dire | This is a synonym for direction. |
horiz | This is a synonym for horizontal. |
horz | This is a synonym for horizontal. |
inde | This is a synonym for indent. |
rota | This is a synonym for rotation. |
shri | This is a synonym for shrink_to_fit. |
shrink | This is a synonym for shrink_to_fit. |
vert | This is a synonym for vertical. |
borders
left | A type of border line* |
right | A type of border line* |
top | A type of border line* |
bottom | A type of border line* |
diag | A type of border line* |
left_colour | A colour. The default is the automatic colour. |
right_colour | A colour. The default is the automatic colour. |
top_colour | A colour. The default is the automatic colour. |
bottom_colour | A colour. The default is the automatic colour. |
diag_colour | A colour. The default is the automatic colour. |
need_diag_1 | A boolean value. The default is False. |
need_diag_2 | A boolean value. The default is False. |
left_color | A synonym for left_colour |
right_color | A synonym for right_colour |
top_color | A synonym for top_colour |
bottom_color | A synonym for bottom_colour |
diag_color | A synonym for diag_colour |
*This can be either an integer width between 0 and 13 or one of the following:
no_line, thin, medium, dashed, dotted, thick, double, hair, medium_dashed, thin_dash_dotted, medium_dash_dotted, thin_dash_dot_dotted, medium_dash_dot_dotted,
slanted_medium_dash_dotted
pattern
back_colour | A colour. The default is the automatic colour. |
fore_colour | A colour. The default is the automatic colour. |
pattern | One of the following: no_fill, none, solid, solid_fill, solid_pattern, fine_dots, alt_bars, sparse_dots, thick_horz_bands, thick_vert_bands, thick_backward_diag, thick_forward_diag, big_spots, bricks, thin_horz_bands, thin_vert_bands, thin_backward_diag, thin_forward_diag, squares, diamonds The default is none. |
fore_color | A synonym for fore_colour |
back_color | A synonym for back_colour |
pattern_fore_colour | A synonym for fore_colour |
pattern_fore_color | A synonym for fore_colour |
pattern_back_colour | A synonym for back_colour |
pattern_back_color | A synonym for back_colour |
protection
The protection features of the Excel file format are only partially implemented in xlwt. Avoid them unless you plan on finishing their implementation.
cell_locked | A boolean value. The default is True. |
formula_hidden | A boolean value. The default is False. |
align
A synonym for alignment border
A synonym for borders
Types of attribute
Boolean values are either True or False, but easyxf allows great flexibility in how you choose to express those two values:
• True can be expressed by 1, yes, true or on
• False can be expressed by 0, no, false, or off
Colours in Excel files are a confusing mess. The safest bet to do is just pick from the following list of colour names that easyxf understands.
The names used are those reported by the Excel 2003 GUI when you are inspecting the default colour palette.
Warning: There are many differences between this implicit mapping from colour-names to RGB values and the mapping used in standards such as HTML andCSS.
aqua black blue blue_gray bright_green brown coral cyan_ega dark_blue dark_blue_ega dark_green dark_green_ega dark_purple dark_red | dark_red_ega dark_teal dark_yellow gold gray_ega gray25 gray40 gray50 gray80 green ice_blue indigo ivory lavender | light_blue light_green light_orange light_turquoise light_yellow lime magenta_ega ocean_blue olive_ega olive_green orange pale_blue periwinkle pink | plum purple_ega red rose sea_green silver_ega sky_blue tan teal teal_ega turquoise violet white yellow |
NB: grey can be used instead of gray wherever it occurs above.
Formatting Rows and Columns
It is possible to specify default formatting for rows and columns within a worksheet. This is done using the set_style method of the Row and Column instances, respectively.
The precedence of styles is as follows:
• the style applied to a cell
• the style applied to a row
• the style applied to a column
It is also possible to hide whole rows and columns by using the hidden attribute of Row and Column instances.
The width of a Column can be controlled by setting its width attribute to an integer where 1 is 1/256 of the width of the zero character, using the first font that occurs in the Excel file. Do not be fooled by the height attribute of the Row class, it does nothing. Specify a style on the row and set its font height attribute instead.
The following example shows these methods and properties in use along with the style precedence:
from xlwt import Workbook, easyxf from xlwt.Utils import rowcol_to_cell row = easyxf('pattern: pattern solid, fore_colour blue') col = easyxf('pattern: pattern solid, fore_colour green') cell = easyxf('pattern: pattern solid, fore_colour red') book = Workbook() sheet = book.add_sheet('Precedence') for i in range(0,10,2): (i).set_style(row) for i in range(0,10,2): (i).set_style(col) for i in range(10): sheet.write(i,i,None,cell) sheet = book.add_sheet('Hiding') for rowx in range(10): for colx in range(10): sheet.write(rowx,colx,rowcol_to_cell(rowx,colx)) for i in range(0,10,2): (i).hidden = True (i).hidden = True sheet = book.add_sheet('Row height and Column width') for i in range(10): sheet.write(0,i,0) for i in range(10): (i).set_style(easyxf('font:height '+str(200*i))) (i).width = 256*i ('') |
Formatting Sheets and Workbooks
There are many possible settings that can be made on Sheets and Workbooks.
Most of them you will never need or want to touch.
If you think you do, see the “Other Properties” section below.
Style compression
While its fine to create as many XFStyle and their associated Font instances as you like, each one written to Workbook will result in an XF record and a Font record. Excel has fixed limits of around 400 Fonts and 4000 XF records so care needs to be taken when generating large Excel files.
To help with this, xlwt.Workbook has an optional style_compression parameter with the following meaning:
• 0 – no compression. This is the default.
• 1 – compress Fonts only. Not very useful.
• 2 – compress Fonts and XF records.
The following example demonstrates these three options:
from xlwt import Workbook, easyxf style1 = easyxf('font: name Times New Roman') style2 = easyxf('font: name Times New Roman') style3 = easyxf('font: name Times New Roman') def write_cells(book): sheet = book.add_sheet('Content') sheet.write(0,0,'A1',style1) sheet.write(0,1,'B1',style2) sheet.write(0,2,'C1',style3) book = Workbook() write_cells(book) ('') book = Workbook(style_compression=1) write_cells(book) ('') book = Workbook(style_compression=2) write_cells(book) ('') |
Be aware that doing this compression involves deeply nested comparison of the XFStyle objects, so may slow down writing of large files where many styles are used.
The recommended best practice is to create all the styles you will need in advance and leave style_compression at its default value.
Formulae
Formulae can be written by xlwt by passing an xlwt.Formula instance to either of the write methods or by using the set_cell_formula method of Row instances, bugs allowing.
The following are supported:
• all the built-in Excel formula functions
• references to other sheets in the same workbook
• access to all the add-in functions in the Analysis Toolpak (ATP)
• comma or semicolon as the argument separator in function calls
• case-insensitive matching of formula names The following are not suppoted:
• references to external workbooks
• array aka Ctrl-Shift-Enter aka CSE formulas
• references to defined Names
• using formulas for data validation or conditional formatting
• evaluation of formulae
The following example shows some of these things in action:
from xlwt import Workbook, Formula book = Workbook() sheet1 = book.add_sheet('Sheet 1') sheet1.write(0,0,10) sheet1.write(0,1,20) sheet1.write(1,0,Formula('A1/B1')) sheet2 = book.add_sheet('Sheet 2') row = (0) row.write(0,Formula('sum(1,2,3)')) row.write(1,Formula('SuM(1;2;3)')) row.write(2,Formula("$A$1+$B$1*SUM('ShEEt 1'!$A$1:$b$2)")) ('') |
Names
Names cannot currently be written by xlwt.
Utility methods
The Utils module of xlwt contains several useful utility functions:
col_by_name
This will convert a string containing a column identifier into an integer column index. cell_to_rowcol
This will convert a string containing an excel cell reference into a four-element tuple containing:
(row,col,row_abs,col_abs)
row – integer row index of the referenced cell col – integer column index of the referenced cell row_abs – boolean indicating whether the row index is absolute (True) or relative (False) col_abs – boolean indicating whether the column index is absolute (True) or relative
(False)
cell_to_rowcol2
This will convert a string containing an excel cell reference into a two-element tuple containing: (row,col)
row – integer row index of the referenced cell col – integer column index of the referenced cell rowcol_to_cell
This will covert an integer row and column index into a string excel cell reference, with either index optionally being absolute.
cellrange_to_rowcol_pair
This will convert a string containing an excel range into a four-element tuple containing:
(row1,col1,row2,col2)
row1 – integer row index of the start of the range col1 – integer column index of the start of the range row2 – integer row index of the end of the range col2 – integer column index of the end of the range rowcol_pair_to_cellrange
This will covert a pair of integer row and column indexes into a string containing an excel cell range. Any of the indexes specified can optionally be made to be absolute. valid_sheet_name
This function takes a single string argument and returns a boolean value indication whether the sheet name will work without problems (True) or will cause complaints from Excel (False).
The following example shows all of these functions in use:
from xlwt import Utils print 'AA ->',Utils.col_by_name('AA') print 'A ->',Utils.col_by_name('A') print 'A1 ->',Utils.cell_to_rowcol('A1') print '$A$1 ->',Utils.cell_to_rowcol('$A$1') print 'A1 ->',Utils.cell_to_rowcol2('A1') print (0,0),'->',Utils.rowcol_to_cell(0,0) print (0,0,False,True),'->', print Utils.rowcol_to_cell(0,0,False,True) print (0,0,True,True),'->', print Utils.rowcol_to_cell( row=0,col=0,row_abs=True,col_abs=True ) print '1:3 ->',Utils.cellrange_to_rowcol_pair('1:3') print 'B:G ->',Utils.cellrange_to_rowcol_pair('B:G') print 'A2:B7 ->',Utils.cellrange_to_rowcol_pair('A2:B7') print 'A1 ->',Utils.cellrange_to_rowcol_pair('A1') print (0,0,100,100),'->', print Utils.rowcol_pair_to_cellrange(0,0,100,100) print (0,0,100,100,True,False,False,False),'->', print Utils.rowcol_pair_to_cellrange( row1=0,col1=0,row2=100,col2=100, row1_abs=True,col1_abs=False, row2_abs=False,col2_abs=True ) for name in ( '',"'quoted'","O'hare","X"*32,"[]:\\?/*\x00" ): print 'Is %r a valid sheet name?' % name, if Utils.valid_sheet_name(name): print "Yes" else: print "No" |
Other properties
There are many other properties that you can set on xlwt-related objects. They are all listed below, for each of the types of object. The names are mostly intuitive but you are warned to experiment thoroughly before attempting to use any of these in an important situation as some properties exist that aren't saved to the resulting Excel files and some others are only
partially implemented.
owner country_code wnd_protect obj_protect protect backup_on_save hpos | vpos width height active_sheet tab_width wnd_visible wnd_mini | hscroll_visible vscroll_visible tabs_visible dates_1904 use_cell_values |
xlwt.Workbook
set_style height has_default_height xlwt.Column | height_mismatch level collapse | hidden space_above space_below |
set_style width_in_pixels | width hidden | level collapse |
xlwt.Worksheet
name visibility row_default_height_mismatch row_default_hidden row_default_space_above row_default_space_below show_formulas show_grid show_headers show_zero_values auto_colour_grid cols_right_to_left show_outline remove_splits selected sheet_visible page_preview first_visible_row first_visible_col grid_colour dialog_sheet auto_style_outline outline_below outline_right fit_num_pages show_row_outline show_col_outline alt_expr_eval alt_formula_entries row_default_height col_default_height calc_mode | save_recalc print_headers print_grid header_str footer_str print_centered_vert print_centered_horz left_margin right_margin top_margin bottom_margin paper_size_code print_scaling start_page_number fit_width_to_pages fit_height_to_pages print_in_rows portrait print_colour print_draft print_notes print_notes_at_end print_omit_errors print_hres header_margin footer_margin copies_num wnd_protect obj_protect protect scen_protect password |
calc_count RC_ref_mode iterations_on delta
Some examples of Other Properties
The following sections contain examples of how to use some of the properties listed above.
Hyperlinks
Hyperlinks are a type of formula as shown in the following example:
from xlwt import Workbook,easyxf,Formula style = easyxf('font: underline single') book = Workbook() sheet = book.add_sheet('Hyperlinks') sheet.write( 0, 0, Formula('HYPERLINK("";"Python")'), style) link = 'HYPERLINK("";"help")' sheet.write( 1,0, Formula(link), style) ("") |
Images
Images can be inserted using the insert_bitmap method of the Sheet class:
from xlwt import Workbook
w = Workbook() ws = w.add_sheet('Image') ws.insert_bitmap('', 0, 0)
w.save('')
NB: Images are not displayed by
Merged cells
Merged groups of cells can be inserted using the write_merge method of the Sheet class:
from xlwt import Workbook,easyxf style = easyxf( 'pattern: pattern solid, fore_colour red;' 'align: vertical center, horizontal center;' ) w = Workbook() ws = w.add_sheet('Merged') ws.write_merge(1,5,1,5,'Merged',style) w.save('') |
Borders
Writing a single cell with borders is simple enough, however applying a border to a group of cells is painful as shown in this example:
from xlwt import Workbook,easyxf tl = easyxf('border: left thick, top thick') t = easyxf('border: top thick') tr = easyxf('border: right thick, top thick') r = easyxf('border: right thick') br = easyxf('border: right thick, bottom thick') b = easyxf('border: bottom thick') bl = easyxf('border: left thick, bottom thick') l = easyxf('border: left thick') w = Workbook() ws = w.add_sheet('Border') ws.write(1,1,style=tl) ws.write(1,2,style=t) ws.write(1,3,style=tr) ws.write(2,3,style=r) ws.write(3,3,style=br) ws.write(3,2,style=b) ws.write(3,1,style=bl) ws.write(2,1,style=l) w.save('') |
NB: Extra care needs to be taken if you're updating an existing Excel file!
Split and Freeze panes
It is fairly straight forward to create frozen panes using xlwt.
The location of the split is specified using the integer vert_split_pos and horz_split_pos properties of the Sheet class.
The first visible cells are specified using the integer vert_split_first_visible and horz_split_first_visible properties of the Sheet class. The following example shows them all in action:
from xlwt import Workbook from xlwt.Utils import rowcol_to_cell w = Workbook() sheet = w.add_sheet('Freeze') sheet.panes_frozen = True sheet.remove_splits = True sheet.vert_split_pos = 2 sheet.horz_split_pos = 10 sheet.vert_split_first_visible = 5 sheet.horz_split_first_visible = 40 for col in range(20): for row in range(80): sheet.write(row,col,rowcol_to_cell(row,col)) w.save('') |
Split panes are a less frequently used feature and their support is less complete in xlwt.
The procedure for creating split panes is exactly the same as for frozen panes except that the panes_frozen attribute of the Worksheet should be set to False instead of True.
However, if you really need split panes, you're advised to see professional help before proceeding!
Outlines
These are a little known and little used feature of the Excel file format that can be very useful when dealing with categorised data.
Their use is best shown by example:
from xlwt import Workbook data = [ ['','','2008','','2009'], ['','','Jan','Feb','Jan','Feb'], ['Company X'], ['','Division A'], ['','',100,200,300,400], ['','Division B'], ['','',100,99,98,50], ['Company Y'], ['','Division A'], ['','',100,100,100,100], ['','Division B'], ['','',100,101,102,103], ] w = Workbook() ws = w.add_sheet('Outlines') for i,row in enumerate(data): for j,cell in enumerate(row): ws.write(i,j,cell) (2).level = 1 (3).level = 2 (4).level = 3 (5).level = 2 (6).level = 3 (7).level = 1 (8).level = 2 (9).level = 3 (10).level = 2 (11).level = 3 (2).level = 1 (3).level = 2 (4).level = 1 (5).level = 2 w.save('') |
Zoom magnification and Page Break Preview
The zoom percentage used when viewing a sheet in normal mode can be controlled by setting the normal_magn attribute of a Sheet instance.
The zoom percentage used when viewing a sheet in page break preview mode can be controlled by setting the preview_magn attribute of a Sheet instance.
A Sheet can also be made to show a page break preview by setting the page_preview attribute of the Sheet instance to True.
Here's an example to show all three in action:
from xlwt import Workbook w = Workbook() ws = w.add_sheet('Normal') ws.write(0,0,'Some text') ws.normal_magn = 75 ws = w.add_sheet('Page Break Preview') ws.write(0,0,'Some text') ws.preview_magn = 150 ws.page_preview = True w.save('') |
Filtering Excel Files
Any examples shown below can be found in the xlutils directory of the course material.
Other utilities in xlutils
The xlutils package contains several utilities in addition to those for filtering. The following are often useful:
xlutils.styles
This module contains one class which, when instantiated with an xlrd.Workbook, will let you discover the style name and information from a given cell in that workbook as shown in the following example:
from xlrd import open_workbook from xlutils.styles import Styles book = open_workbook('',formatting_info=True) styles = Styles(book) sheet = book.sheet_by_index(0) print styles[(1,1)].name print styles[(1,2)].name A1_style = styles[(0,0)] A1_font = book.font_list[.font_index] print book.colour_map[A1_font.colour_index] |
NB: For obvious reasons, open_workbook must be called with formatting_info=True in order to use xlutils.styles.
Full documentation and examples can be found in the file in the docs folder of xlutils' source distribution.
xlutils.display
This module contains utility functions for easy and safe display of information returned by xlrd.
quoted_sheet_name is called with the name attribute of an xlrd.sheet.Sheet instance and will return an encoded string containing a quoted version of the sheet's name.
cell_display is called with an instance and returns an encoded string containing a sensible representation of the cells contents, even for Date and Error cells. If a date cell is to be displayed, cell_display must be called with the datemode attribute of the from which the cell came.
The following examples show both functions in action:
from xlrd import open_workbook from xlutils.display import quoted_sheet_name from xlutils.display import cell_display wb = open_workbook('') print quoted_sheet_name(wb.sheet_names()[0]) print repr(quoted_sheet_name(u'Price(\xa3)','utf-8')) print quoted_sheet_name(u'My Sheet') print quoted_sheet_name(u"John's Sheet") sheet = wb.sheet_by_index(0) print cell_display((1,1)) print cell_display((1,3),wb.datemode) |
Full documentation and examples can be found in the file in the docs folder of xlutils' source distribution.
This module contains one function that will take an and returns an xlwt.Workbook populated with the data and formatting found in the .
This is extremely useful for updating an existing spreadsheet as the following example shows:
from xlrd import open_workbook from xlwt import easyxf from import copy rb = open_workbook('',formatting_info=True) rs = rb.sheet_by_index(0) wb = copy(rb) ws = wb.get_sheet(0) plain = easyxf('') for i,cell in enumerate((2)): if not i: continue ws.write(i,2,cell.value,plain) for i,cell in enumerate((4)): if not i: continue ws.write(i,4,cell.value-1000) ('') |
It is important to note that some things won't be copied:
• Formulae
• Names
• anything ignored by xlrd
In addition to the modules described above, there are also xlutils.margins and , but these are only useful in certain situations. Refer to their documentation in the xlutils source distribution.
Structure of xlutils.filter
This framework is designed to filter and split Excel files using a series of modular readers, filters and writers as shown in the diagram below:
The flow of information between the components is by method calls on the next component in the chain. The possible method calls are listed in the table below, where rdbook is an instance, rdsheet is an xlrd.sheet.Sheet instance, rdrowx, rdcolx, wtrowx and wtcolx and integer indexes specifying the cell to read from and write to, wtbook_name is a string specifying the name of the Excel file to write to and wtsheet_name is a unicode specifying the name of the sheet to write to:
start() | This method is called before processing of a batch of input. It can be called at any time. One common use is to reset all the filters in a chain in the event of an error during the processing of an rdbook. |
workbook(rdbook,wtbook_name) | This method is called every time processing of a new workbook starts |
sheet(rdsheet,wtsheet_name) | This method is called every time processing of a new sheet in the current workbook starts |
set_rdsheet(rdsheet) | This method is called to indicate a change for the source of cells mid-way through writing a sheet. |
row(rdrowx,wtrowx) | The row method is called every time processing of a new row in the current sheet starts. |
cell(rdrowx,rdcolx,wtrowx,wtcolx) | This is called for every cell in the sheet being processed. This is the most common method in which filtering and queuing of onward calls to the next component takes place. |
finish | This method is called once processing of all workbooks has been completed. |
Readers
A reader's job is to obtain one or more objects and iterate over those objects issuing appropriate calls to the next component in the chain. The order of calling is expected to be as follows:
• start
◦ workbook, once for each object obtained
▪ sheet, once for each sheet found in the current book
▪ set_rdsheet, whenever the sheet from which cells to be read needs to be changed. This method may not be called between calls to row and cell, and between multiple calls to cell. It may only be called once all cell calls for a row have been made.
• row, once for each row in the current sheet
◦ cell, once for each cell in the row
• finish, once all objects have been processed
Also, for method calls made by a reader, the following should be true:
• wtbook_name should be the filename of the file the object originated from.
• wtsheet_name should be
• wtrowx should be equal to rdrowx
• rdcolx should be equal to wtcolx
Because of these restrictions, an xlutils.filter.BaseReader class is provided that will normally only need to have one of two methods overridden to get any required functionality:
• get_filepaths – if implemented, this must return an iterable sequence of paths to excel files that can be opened with python's builtin file.
• get_workbooks – if implemented, this must return an sequence of 2-tuples. Each tuple must contain an object followed by a string containing the filename of the file from which the object was loaded.
Filters
Implementing these components is where the bulk of the work will be done by users of the xlutils.filter framework. A Filter's responsibilities are to accept method calls from the preceding component in the chain, do any processing necessary and then emit appropriate method calls to the next component in the chain.
There is very little constraint on what order Filters receive and emit method calls other than that the order of method calls emitted must remain consistent with the structure given above. This enables components to be freely interchanged more easily.
Because Filters may only need to implement few of the full set of method calls, an xlutils.filter.BaseFilter is provided that does nothing but pass the method calls on to the next component in the chain. The implementation of this filter is useful to see when embarking on Filter implementation:
class BaseFilter: def start(self): .start() def workbook(self,rdbook,wtbook_name): .workbook(rdbook,wtbook_name) def sheet(self,rdsheet,wtsheet_name): self.rdsheet = rdsheet .sheet(rdsheet,wtsheet_name) def set_rdsheet(self,rdsheet): self.rdsheet = rdsheet .set_rdsheet(rdsheet) def row(self,rdrowx,wtrowx): (rdrowx,wtrowx) def cell(self,rdrowx,rdcolx,wtrowx,wtcolx): (rdrowx,rdcolx,wtrowx,wtcolx) def finish(self): .finish() |
Writers
These components do the grunt work of actually copying the appropriate information from the rdbook and serialising it into an Excel file. This is a complicated process and not for the feint of hard to re-implement.
For this reason, an xlutils.filter.BaseWriter component is provided that does all of the hard work and has one method that needs to be implemented. That method is get_stream and it is called with the filename of the Excel file to be written.
Implementations of this method are expected to return a new file-like object that has a write and, by default, a close method each time they are called.
Subclasses may also override the boolean close_after_write attribute, which is True by default, to indicate that the file-like objects returned from get_stream should not have their close method called once serialisation of the Excel file data is complete.
It is important to note that some things won't be copied from the rdbook by BaseWriter:
• Formulae
• Names
• anything ignored by xlrd
Process
The process function is responsible for taking a series of components as its arguments. The first of these should be a Reader. The last of these should be a Writer. The rest should be the necessary Filters in the order of processing required.
The process method will wire these components together by way of their next attributes and then kick the process off by calling the Reader and passing the first Filter in the chain as its argument.
A worked example
Suppose we want to filter an existing Excel file to omit rows that have an X in the first column.
The following example shows possible components to do this and shows how they would be instantiated and called to achieve this:
import os from xlutils.filter import \ BaseReader,BaseFilter,BaseWriter,process class Reader(BaseReader): def get_filepaths(self): return [.abspath('')] class Writer(BaseWriter): def get_stream(self,filename): return file(filename,'wb') class Filter(BaseFilter): pending_row = None wtrowxi = 0 def workbook(self,rdbook,wtbook_name): .workbook(rdbook,'filtered-'+wtbook_name) def row(self,rdrowx,wtrowx): self.pending_row = (rdrowx,wtrowx) def cell(self,rdrowx,rdcolx,wtrowx,wtcolx): if rdcolx==0: value = (rdrowx,rdcolx).value if value.strip().lower()=='x': self.ignore_row = True self.wtrowxi -= 1 else: self.ignore_row = False rdrowx, wtrowx = self.pending_row (rdrowx,wtrowx+self.wtrowxi) elif not self.ignore_row: ( rdrowx,rdcolx,wtrowx+self.wtrowxi,wtcolx-1 ) process(Reader(),Filter(),Writer()) |
In reality, we would not need to implement the Reader and Writer components, as there are already suitable components included.
Existing components
The xlutils.filter framework comes with a wide range of existing components, each of which is briefly described below. For full descriptions and worked examples of all these components, please see in the docs folder of the xlutils source distribution.
GlobReader
If you're processing files that are on disk, then this is probably the reader for you. It returns all files matching the path specification it's instantiated with.
XLRDReader
This reader can be used at the start of a chain when you already have an object and you'll looking to process it with xlutils.filter.
TestReader
This reader is specifically designed for testing filterimplementations with known sets of cells.
DirectoryWriter
If you want files you're processing to end up on disk, then this is probably the writer for you. It stores files in the directory it is instantiated with.
StreamWriter
If you want to write exactly one workbook to a stream, such as a tempfile.TemporaryFile or sys.stdout, then this is the writer for you.
XLWTWriter
If you want to change cells after the filtering process is complete then this writer can be used to obtain the xlwt.Workbook objects that BaseWriter generates.
ColumnTrimmer
This filter will strip columns containing no useful data from the end of sheets. The definition of “no useful data” can be controlled during instantiation of this filter.
ErrorFilter
This filter caches all method calls in a file on disk and will only pass them on the next component in the chain when its finish method has been called and no error messages have been logged to the python logging framework.
If Boolean or error Cells are encountered, an error message will be logged to the python logging framework will will also usually mean that no methods will be emitted from this component to the next component in the chain.
Finally, cell method calls corresponding to Empty cells in rdsheet will not be passed on to the next component in the chain.
Calling this component's start method will reset it.
Echo
This filter will print calls to the methods configured when the filter is instantiated along with the arguments passed.
MemoryLogger
This filter will dump stats to the path it was configured with using the heapy package if it is available. If it is not available, no operations are performed.
For more information on heapy, please see
Possible Tasks for Workshop
The following is a list of tasks that can be attempted by any attendee who hasn't brought their own tasks to attempt.
Installation with IronPython
The libraries have been used successfully with IronPython, but this has not been thoroughly tests or documented.
Installation with Jython
The libraries should all work with Jython, but no one has so far attempted to do so.
Inserting a row into a sheet
Starting with an existing Excel file, attempt to create a new Excel file with a row inserted at a given position.
Splitting a Book into its Sheets
Starting with an existing Excel file, create a directory containing one file for each worksheet in the original file.
Reporting errors in a directory full on Excel files
Scan a directory of Excel files and report the location of any error cells.
A progression of this task is to allow the passing of options to indicate what types of error to report.
Removing Rows containing errors
Starting with an existing Excel file, create a filtering process that generates a new Excel file that excludes any rows containing error cells.
A progression of this task is to generate a new Excel file that contains empty cells where there were errors in the original file.
Filtering Excel files to and from a web server
This task is to create components for xlutils.filter that can read from a website and write back to that website.
The task should result in an HTTPReader and an HTTPWriter.
Producing a report from a database
This task is to take a typical database query and dump it into an Excel file such that the heading row is set up nicely with decent alignment in a frozen pane.
As a precursor to this task, you may need to set up a typical database!