Skip to contents

Changelog

Source: NEWS.md

openxlsx2 1.15

CRAN release: 2025-04-25

Maintenance

  • Legacy create_comment() is now deprecated in favor of wb_comment().
  • Usage of rows and cols is now deprecated in favor of wb_dims(rows = ..., cols = ...).

New features

  • A normalization function was added for separators in formulas. In English locales, formulas require a comma as argument separator, locales that use the comma a decimal separator, a semicolon is used as argument separator. We added a check to replace semicolons added in formulas, because these were hard to spot and if overlooked could brick the XML worksheet.

Fixes

  • Grouping of rows does no longer require the rows to be initialized in the worksheet. 1303
  • Fix another case where cc was unintentionally shorten when applying a cm formula. Not sure if the cause for this is in the cm formula or in some other place.
  • write_xlsx() now handles table_name as documented. 1314
  • Writing the character strings "Inf", "-Inf", and "NaN" is now possible. 1317

Internal changes

  • Previously, when loading a workbook, the included styles were broken down, with XML styles converted into data frame objects. However, these results were only used to generate a sequence.
  • Similarly, if available, the entire shared strings table was converted to text upon loading. Now, this conversion is postponed until explicitly requested. Both approaches have their advantages: reading shared strings on load is useful when they are extensively shared across worksheets and multiple worksheets are retrieved via wb_to_df(), while delaying conversion saves memory and computation time when not required. However, in the latter case, shared strings across multiple worksheets will need to be converted each time they are accessed. The impact of this change should be monitored.
  • dims_to_dataframe() was improved to better handle many combined dims like “A1:B1,A3:D4”

Breaking changes

  • Improved performance when reading .xlsx files containing a large number of datetime variables. Datetime columns are now consistently returned as POSIXct objects in the UTC timezone. 1331
    • Reason: Standardizing datetime representation to UTC simplifies internal handling and eliminates complexities related to local timezones and Daylight Saving Time during data import, leading to performance benefits.
    • Impact: Previously, datetimes were returned in the system’s local timezone (Sys.timezone()). Users will now receive datetimes in UTC (same date and time, but different timezone). If local time representation is needed, explicit conversion using functions like lubridate::force_tz() is required after reading the data.

openxlsx2 1.14

CRAN release: 2025-03-20

New features

  • A new experimental flush argument has been introduced to wb_save(), allowing a custom XML streaming function for worksheets to help prevent memory spikes. This feature has only been tested within openxlsx2 and not extensively with spreadsheet software. Since it bypasses certain fail-safe mechanisms, including XML validity checks, it should only be used as a last-resort solution. 1255 This expects all inputs to be UTF-8, therefore it might not work with R running in different locale sessions, if non ASCII characters are used.

Fixes

  • Input validation has been added to fmt_txt(), similar to how it has been added to the create_*() family a while ago. 1280
  • Aligned with the standard, openxlsx2 now uses a baseColWith of 8 (previously 8.43). The standard requires an integer and we provided a float. This fixes an issue with third party software. 1284 Files with the previously incorrect baseColWidth can be loaded and saved and the fix will be applied.
  • Fixed a bug, where it was impossible to add data to a worksheet with a cm formula. This is a regression that was introduced with the internal changes in the previous release. The fix also restores writing cm formulas. 1287
  • Some changes were implemented to ease interaction with the python library openpyxl used by pandas. 1289, 1291
  • Corrected the documentation for wb_add_font(). 1293

Breaking changes

  • Single vector character classes will be treated as vector input. This extends the handling of glue, added in the previous release, and covers now all character classes, including fmt_txt().

openxlsx2 1.13

CRAN release: 2025-02-15

New features

Fixes

  • The first formula in a workbook can now be a shared formula. 1223
  • Avoid passing ASCII strings through Rcpp::String(). Previously all cc columns were passed through Rcpp::String() to avoid encoding issues on non unicode systems. 1224
  • fmt_txt() is now indifferent about color and colour. 1229
  • Improve set_col_widths(widths = "auto"). This should avoid very wide columns for numeric columns. 1239

Internal changes

  • Update of internal pugixml library
  • Switch to f_attr to handle more formula attributes
  • Remove the use of cc_out when writing output files
  • Refactoring of wb_add_border() (@pteridin)
  • Trim internal cc data frame to only a selected set of columns

Breaking changes

  • Style helpers will accept colors only if provided via wb_color(). This broke a single example case that was still using c(rgb = "FF808080") this can be converted to wb_color(hex = "FF808080").

openxlsx2 1.12

CRAN release: 2024-12-19

New features

Fixes

  • Create date is not reset to the present time in each call to wb_set_properties(). 1176
  • Improve handling of file headers and footers for a case where wb_load() would previously fail. 1186
  • Partial labels were written only over the first element and only if assigned in an ordered fashion. 1189
  • Enable use of current_sheet() in create_sparklines(). This is the default for the function, but was not supported.
  • When setting headers and footers, elements can now be skipped with NA as documented. 1211

Breaking changes

  • wb_workbook() and the underlying wbWorkbook object gained a new argument datetime_modified. This argument was added after datetime_created and changes the number and ordering of arguments. 1176

openxlsx2 1.11

CRAN release: 2024-11-14

New features

Fixes

  • Previously rows that trigger scientific notation (e.g. 1e+05) would cause issues, when matched against a non scientific version. 1170
  • When using wb_add_data_table(..., total_row = TRUE) the last row of the data table in the workbook was mistakenly overwritten with the total row formula, which should have been placed below the last row of the table. This caused loss of data. 1179

openxlsx2 1.10

CRAN release: 2024-10-17

New features

  • When writing a file with na.strings = NULL, the file will not contain any reference to the blank cell. Depending on the number of missings in a data set, this can reduce the file size significantly. 1111

  • wb_to_df() gained a new argument show_hyperlinks which returns the target or location of a hyperlink, instead of the links description. 1136

  • A new wrapper function wb_add_hyperlink() extends the capabilities of writing hyperlinks to shared hyperlinks. Shared hyperlinks bring along internal changes that are noted below. 1137

  • Add address field to wb_add_image(). This can be used to add a url or mailto address to an image (1138, conversion from a PR by @jistria for openxlsx)

  • It is now possible to remove hyperlinks with either wb_remove_hyperlinks() or wb_clean_sheet(). 1139

  • wb_add_data_table() will throw a warning if non distinct column names are found and will fix this for the user. Non distinct can be duplicated or even upper and lower case x and X. 1150

Fixes

  • The integration of the shared formula feature in the previous release broke the silent extension of dims, if a single cell dims was provided for an x that was larger than a single cell in wb_add_formula(). 1131

  • Fixed a regression in the previous release, where wb_dims() would pass column names passed via cols to col2int() which could cause overflow errors resulting in a failing check. 1133

  • Fix cloning from worksheets with multiple images.

  • Improved wb_to_df(types = ...). Previously if used on unordered data this could cause unintended class order. types no longer requires knowledge of the order of the variables. Not all variables must be specified and it accepts character classes as well. 1147

  • Improve creation of table ids. Previously, on a dirty workbook, unique table ids were not enforced. This caused duplicated table ids, which lead to errors in spreadsheet software. 1152

Internal changes

  • The handling of shared hyperlinks has been updated. Previously, when loading a file with shared hyperlinks, they were converted into wbHyperlink objects (a legacy from openxlsx). With recent internal changes, hyperlinks are no longer automatically transformed into wbHyperlink objects. If you still require these objects, you can use the internal function wb_to_hyperlink(wb, sheet = 1). However, please note that this class is not essential for openxlsx2 and may be further simplified or removed in the future without notice. 1137

openxlsx2 1.9

CRAN release: 2024-09-08

New features

  • Experimental support for shared formulas. Similar to spreadsheet software, when a cell is dragged to horizontally or vertically. This requires the formula to be written only for a single cell and it is filled by spreadsheet software for the remaining dimensions. wb_add_formula() gained a new argument shared. 1074

  • Experimental support for reading shared formulas. If show_formula is used with wb_to_df(), we try to show the value that is shown in spreadsheet software. 1091

  • It is possible to read cells containing formulas as formula. 1103

Fixes

  • If a font unknown to openxlsx2 is used wb_set_col_widths() defaults to using the workbooks default font. 1080

  • Previously, if only cols and rows were passed to wb_dims() and row 1 was selected, incorrect results were returned. This has been fixed. 1094

  • wb_dims() no longer ignores above, below, left, right if from_dims is not supplied. 1104

  • wb_to_df() with skip_hidden_rows = TRUE works now if a file path is passed. 1122

openxlsx2 1.8

CRAN release: 2024-07-07

Maintenance

New features

  • wb_dims() is now able to handle various columns. 1019

  • wb_to_df() now has a check_names argument. 1050

  • The set of conditional formatting icon sets now includes x14 icons. This commit also fixed adding conditional formatting to worksheets with pivot tables. 1053

Fixes

  • Many improvements in the xlsb parser. This includes changes to the logic of the formula parser, rich text strings are now handled, data validation, table formulas and various corrections all over the place. It is still lacking various features and this wont change in the foreseeable future, but the parser is now in better shape than ever. 1037, 1040, 1042, 1044, 1049, 1054

  • write_xlsx() now uses sheet. Previously it required the undocumented sheet_name. 1057

  • Fixed a bug were we obfuscated valid html in worksheets with vml buttons. These files previously did not load. 1062

  • Fixed slow writing of non consecutive number formats introduced in the previous release. 1067, 1068

openxlsx2 1.7

CRAN release: 2024-06-02

New features

Fixes

  • Fixed an issue with non consecutive dims, where columns or rows were silently dropped. 1015

  • Fixes to wb_clone_worksheet() cloning drawings and images should be restored. 1016

  • Fixed an issue where non consecutive columns with special types would overlap. If columns A and C were dates, column B would be formatted as date too. 1026

openxlsx2 1.6

CRAN release: 2024-05-01

New features

  • Helper to read sensitivity labels from files and apply them to workbooks. 983

  • It is now possible to pass non consecutive dims like "A1:B1,C2:D2" to various style helpers like wb_add_fill(). In addition it is now possible to write a data set into a predefined dims region using enforce = TRUE. This handles either "," or ";" as cell separator. 993

# force a dataset into a specific cell dimension
wb <- wb_workbook()$add_worksheet()
wb$add_data(dims = "I2:J2;A1:B2;G5:H6", x = matrix(1:8, 4, 2), enforce = TRUE)

Fixes

  • Allow writing data frames with zero rows. 987

  • wb_dims() has been improved and is safer on 0-length inputs. In particular, it will error for a case where a cols doesn’t exist in x (990, @olivroy).

# Previously created a wrong dims
wb_dims(x = mtcars, cols = "non-existent-col")
# Now errors
  • wb_set_col_widths() is more strict about its arguments. If you provide cols, widths, or hidden don’t have appropriate length, it will throw a warning. This may change to an error in the future, so it is recommended to use appropriate values. (991, @olivroy).

openxlsx2 1.5

CRAN release: 2024-03-26

(This was updated post release.)

New features

  • It’s now possible to pass array formula vectors to wb_add_formula(). 958

  • wb_add_data_table() gained a new total_row argument. This allows to add a total row to spreadsheets including text and spreadsheet formulas. 959

  • wb_dims() now accepts from_dims to specify a starting cell 960.

  • You can now set options(openxlsx2.na.strings) to a value to have a default value for na.strings in wb_add_data(), wb_add_data_table(), and write_xlsx() 968.

  • The direction vectors are written is now controlled via dims. Previously it was required to transpose a vector to write it horizontally: wb_add_data(x = t(letters), col_names = FALSE). Now the direction is defined by dims. The default is still to write vectors vertically, but for a horizontal vector it is possible to write wb_add_data(x = letters, dims = "A1:Z1"). This change impacts vectors, hyperlinks and formulas and basically everything that is not a two dimensional x object.

# before (workaround needed)
wb$add_data(dims = wb_dims(rows = 1, cols = 1:3), x = t(c(4, 5, 8)), col_names = FALSE)
# now (listens to dims)
wb$add_data(dims = wb_dims(rows = 1, cols = 1:3), x = c(4, 5, 8))

Fixes

openxlsx2 1.4

CRAN release: 2024-02-15

New features

  • Experimental support for calculation of pivot table fields. 892

  • Improve sparkline creation with new options and support snake case arguments. 920

  • Experimental support to get and set base colors. 938

Fixes

  • Character strings with XML content were not written correctly: a <br/> b was converted to something neither we nor spreadsheet software was able to decipher. 895

  • Restore first_active_row/first_active_col and first_col/first_row functionality in write_xlsx(). 903

  • Further attempts to fix pivot table sorting. 912 In addition improve handling of non distinct names in wb_data() and add create_pivottablestyle(). 914

  • Document adding background color and images to comments and fix adding more than two images as background. 919

  • Improvements to wb_set_base_font(). This modifies the workbook theme now, including panose values. 935

  • Hyperlinks now use the color of the theme and the base size. 937

  • wb_add_data() and wb_add_data_table() yield better error messages if attempting to add data to an empty workbook (942, @olivroy).

Breaking changes

  • Updating to themes. This includes updates to the default style 'Office Theme' 899
    • This includes switching to the new default font 'Aptos Narrow'
    • A new style 'Office 2013 - 2022 Theme' was added
    Users that want to remain on the old style should use wb_workbook(theme = 'Office 2013 - 2022 Theme') or wb_set_base_font(font_name = "Calibri").

Maintenance

openxlsx2 1.3

CRAN release: 2024-01-12

Documentation improvement

  • Further tweaks to documentation and vignettes to make them more consistent.
    • wb_add_pivot_table() / wb_add_slicer()
    • wb_load(): calc_chain is no longer visible and the previous text that might have been misleading in regards of its use, has been replaced by a more detailed description of what are the consequences of keeping the calculation chain

New features

  • Allow further modifications of comments. The background can now be filled with a color or an image. 870

  • Added wb_set_cell_style_across() to apply a cell style to selected columns and rows. This allows unlocking cells to make use column and row properties of wb_protect_worksheet() which require additional cell styles (see issue 871 for a more detailed explanation). wb_set_cell_style() now accepts a cell dimension in the style argument. 873

Fixes

  • wb_add_ignore_error() now returns a wbWorkbook. 865

  • Deactivate the is_hyperlink check for non-dataframe objects in wb_add_data(). Internally, vapply() is applied to the input object, which is applied column-wise for a data frame and cell-wise for a matrix. This speeds up the writing of larger matrices considerably. 876

  • Column style currency is now correctly applied to numeric vectors. Previously it was not handled. This applies the built in spreadsheet style for currency presumably linked to the spreadsheet software locale. 879

  • wb_to_df(col_names = FALSE) no longer drops column names from logical vectors. Previously, column names were replaced by NA. Now the column name is returned as a cell value in a character column. 884

openxlsx2 1.2

CRAN release: 2023-11-29

New features

  • Add new params to wb_add_pivot_table(). It is now possible to set the show_data_as value and set a tabular table design. 833

Fixes

  • Previously formulas written as data frames were not xml escaped. 834

  • Improve drawing relationship id selection that could cause issues with unordered relationship ids in loaded workbooks. 838

  • Improve copying cells in transpose mode and with hyperlinks. 850

  • Options openxlsx2.maxWidth and openxlsx2.minWidth are now respected as documented when setting column widths with wb_set_col_widths(). 847

openxlsx2 1.1

CRAN release: 2023-10-23

New features

  • Basic support for cloning worksheets across workbooks was added in wb_clone_worksheet(). This should work with tabular and formula data, images and charts. Support for more complex worksheets is known pending. 622

  • Enable loading workbooks without sheets with wb_load(sheet = NULL). This is helpful if the sheets contain a lot of data and only some general workbook data is of interest. 810

  • It is no longer needed to manually create columns for wb_group_cols(). 781

  • It is now possible to sort pivot tables created by wb_add_pivot_table(). 795

  • Basic support for the xlsb file format. We parse the binary file format into pseudo-openxml files that we can import. Therefore, after importing, it is possible to interact with the file as if it had been provided as xlsx in the first place. This is of course slower than reading directly from the binary file. Our implementation is also still missing some features: some array formulas are still broken, conditional formatting and data validation are not implemented, nor are pivot tables and slicers. 688

    • Please note that openxlsx2 is not a security tool and the xlsb parser was written with the intention of reading valid xlsb files on little endian systems.
    • Please do not raise issues about xlsb in terms of speed or completeness of features of the openxml standard. If you have access to other file conversion tools, such as spreadsheet software, they may provide a better solution for your needs.
    • Writing xlsb files is outside the scope of this project.

  • New set of function wb_get_properties()/wb_set_properties() to view and modify workbook properties. 782 This was subsequently improved to handle more workbook properties like company and manager. (799, @olivroy)

  • Basic (experimental) support to add slicers to pivot tables created by openxlsx2. 822

Fixes

  • Removing the worksheet that is the active tab does no longer result in warnings in spreadsheet software. 792

  • update_table() works now without if the table has no autofilter. 802

openxlsx2 1.0

CRAN release: 2023-08-25

Breaking changes

Deprecated functions

These functions are no longer recommended. A guide was created to help users They will continue to work for some time, but changing to newer functions is recommended.

New features

  • wb_comment() is a new helper function to help create wbComment objects It is very similar to create_comment(), with the following differences:

    • author looks at options("openxlsx2.creator"); (create_comment() only used sys.getenv("USERNAME") / Sys.getenv("USER"))
    • visible defaults to FALSE to account for modern spreadsheet software behavior. (create_comment(), it is TRUE).
    • width and height must now be of length 1. (In create_comment(), the first element is taken, other are ignored.)

  • wb_get_sheet_names() gains a escape argument to allow special XML characters to be escaped. 252

  • wb_color() now accepts hex colors with leading sharp (e.g. “#FFFFF”) 728.

  • wb_merge_cells() gains a solve argument. This allows to solve cell intersecting regions. 733

  • wb_add_comment(comment = "x") no longer errors when a comment as a character vector no longer fails (758, @olivroy)

Documentation improvement

  • Tweaks to documentation and vignettes to make them more consistent.

Fixes

  • wb_to_df() now handles date1904 detection. Previous results from this somewhat rare file type were using the default timezone origin of 1900-01-01. 737
  • wb_load() handles more cases.
    • With workbooks with threaded comments 731
    • With workbooks with embeddings other than docx 732
    • with workbooks with long hyperlinks 753
  • wb_load() adds the path to the wbWorkbook object. 741
  • wb_set_header_footer() now works with special characters 747

Internal changes

openxlsx2 0.8

CRAN release: 2023-08-07

API Change

Breaking changes

  • Order of arguments in wb_add_named_region() changed, because previously overlooked dims argument was added.

  • In various functions the order of dims was changed, to highlight it’s importance

  • Cleanups

New features

  • wb_dims() was added as a more convenient replacement for rowcol_to_dims().(691 and 702, @olivroy) The new function can take either numeric (for rows or columns) or character (column) vectors, in addition it is able to create dimensions for R objects that are coercible to data frame. This allows the following variants:
    • wb_dims(1:5, letters)
    • wb_dims(1:5, 1:26)
    • wb_dims(x = matrix(1, 5, 26))
    • wb_dims(x = mtcars, from_col = "C", from_row = 2, row_names = TRUE)
  • Handling of thread comments is now possible via wb_add_thread(). This includes options to reply and resolve comments.

Fixes

  • Improve show_formula. Previously it was called to early in the function and skipped a few cases. 715

Refactoring

openxlsx2 0.7.1

CRAN release: 2023-06-30

New features

  • It is now possible to apply a specific theme to a workbook. 630

  • Removed a few of the former example files and updated the code to use a new default example. This changes internal testing to only run locally if online and external files are required. This reduces the package footprint a little, because 1MB of xlsx files are now excluded. 632

  • The handling of fmt_txt() objects has been improved. It now creates objects of class fmt_txt with their own print(), +, and as.character() methods. Such objects can now also be used as text in create_comment(). 636

  • Improve support for inputs with labels attribute. If e.g. a factor label is numeric, we now try to write the label as number. This should impact the way other partially labeled variables are written. 639

  • Added new wrapper function wb_add_named_style() this supports pre-defined theme aware cell styles like Title or Note. In addition loading of cell styles was improved and additional custom cell styles should be available as well. 628

  • Provide additional options to write special characters in non-unicode environments. 641

  • Add wb_add_dxfs_style() as single line wrapper to create dxf styles used in conditional formatting and custom table styles. 665

Fixes

  • On load app.xml is now assigned to wb$app. Previously it was loaded but not assigned. 629

  • Previously if wb_to_df() was used with argument cols, columns that were missing were created at the end of the output frame. Now columns are returned ordered. 631

  • Fix a bug in wb_load() that was modifying the cell range of conditional formatting. 647

Breaking changes

  • Order of arguments in wb_add_conditional_formatting() changed, because previously overlooked dims argument was added. 642

  • New argument gradientFill was added to create_dxfs_style(). 651

  • Special characters are now escaped in conditional formatting. Hence, previously manually escaped conditional formatting needs updates. 666

openxlsx2 0.7

CRAN release: 2023-05-26

New features

  • The dims argument of wb_add_formula() can be used to create a array references. A new cm argument was added which might be useful, if formulas previously created addition @ in spreadsheet software. Examples how to use formulas were added to a new vignette. 593

  • Allow using custom data table styles. This fixes a few minor style inconsistencies. 594

  • Allow reading and writing hms columns. 601

  • Import tableStyles with wb_load() and improve dxf style creation. 603

  • Add fmt_txt() to style character strings. 607

  • Add new wrapper to ignore worksheet errors wb_add_ignore_error(). 617

  • Add new wrapper to update table references wb_update_table(). 606

Fixes

  • Improve handling of non standard OutDec options. 611

  • openxlsx2 now does a better job of trying to return character values from classes that are foreign to it. This has been going on for quite some time, although previously we had a bug that treated such classes as numeric, resulting in corrupted xlsx files. 615

  • We now return a few additional xml arguments from worksheets. 617

openxlsx2 0.6.1

CRAN release: 2023-04-21

New features

  • Improve col2int() to accept column ranges like col2int("A:Z"). This should allow using column ranges in various places like wb_merge_cells(cols = "B:D", ...) or wb_read(cols = c("A","C:D")). 575

  • Add dims argument to wb_add_image() and wb_add_plot(). This can be used to place images starting at a cell or span a cell range. This deprecates xy in wb_add_plot(). This adds colOffset and rowOffset to wb_add_drawing() and wb_add_mschart() and wb_add_chart_xml(). 578

  • Add skipHiddenCols and skipHiddenRows to wb_to_df(). In this way, hidden columns and rows are ignored, assuming that the person who has hidden them assumes that they are not important. 579

  • When writing tibble use to.data.frame() just like in the data.table case. 582

  • Add cleanup internal comment code in write_comment(). This should not impact the workbook wrapper code in wb_add_comment(). 586

  • Added chain functions for wb_to_df() and wb_load(). 587

openxlsx2 0.6

CRAN release: 2023-04-02

New features

  • Styles arguments now accept logical and numeric arguments where applicable. 558

  • Adding dims argument to wb_clean_sheet(). This allows to clean only a selected range. 563

Fixes

  • na.strings = NULL is no longer ignored in write_xlsx() 552

  • Explicit type conversion to date and datetime is finally available. 551

Breaking changes

  • skipEmptyCols and skipEmptyRows behavior in wb_to_df() related functions was switched to include empty columns that have a name. Previously we would exclude columns if they were empty, even if they had a name. 555

  • Cleanups in 548

    • remove deprecated functions
      • cloneSheetStyle()
      • get_cell_style()
      • set_cell_style()
      • wb_conditional_formatting()
    • remove deprecated arguments
      • xy argument for write_data_table() interacting functions
      • file from xl_open()
      • definedName from wb_to_df() interacting functions
    • deprecating function

openxlsx2 0.5.1

CRAN release: 2023-02-26

New features

  • Allow hierarchical grouping. wb_group_cols/wb_group_rows now accept nested lists as grouping variable. 537

  • It is now possible to add form control types Checkbox, Radio and Drop to a workbook using wb_add_form_control(). 533

  • Improve wb_to_df(fillMergedCells = TRUE) to work better with dimensions. It is now possible to fill cells where the merged cells intersect with the selected dimensions. 541

  • Speedup cell initialization. This is used in wb_style functions like wb_add_numfmt(). The previous loop was replaced with a faster implementation. 545

  • Improve date detection in wb_to_df(). This improves date and posix detection with custom date formats. 547

  • na_strings() is now used as the explicit default value for na.strings parameters in exported workbook functions. 473

  • waiver functions (i.e., next_worksheet(), current_worksheet(), and na_strings()) are now exports 474

Fixes

  • Fixed a bug when loading input with multiple sheets where not every sheet contains a drawing/comment. Previously we assumed that every sheet had a comment and ordered them incorrectly. This caused confusion in spreadsheet software. 536

  • Fixed a bug with files containing 10 or more external references. In this case we did not load the references in numeric order and instead as “1.xml”, “10.xml”, …, “2.xml”, … This jumbled up the external references. 538

openxlsx2 0.5

CRAN release: 2023-01-29

New features

  • Improve column and row grouping. It is now possible to group by list, so that you can create various levels of groupings. 486

  • writeData() calls force(x) to evaluate the object before options are set (#264)

  • tabColor in wb_add_worksheet() now allows passing wb_color(). 500

  • Add wb_copy_cells() a wrapper that allows copying cell ranges in a workbook as direct copy, as reference or as value. 515

  • Experimental option: openxlsx2.string_nums to write string numerics differently. A string numeric is a numeric in a string like: as.character(1.5). The option can be

    • 0 = the current default. Writes string numeric as string (the incorrect way according to spreadsheet software)
    • 1 = writes string numeric as numeric with a character flag (the correct way according to spreadsheet software)
    • 2 = convert all string numeric to numeric when writing

    This is experimental, because the impact is somewhat unknown. It might trigger unintended side effects. Feedback is requested.

  • Enable writing strings as sharedStrings with argument inline_strings = FALSE. This creates a sharedStrings table in openxml that allows to reuse strings in the workbook efficiently and can reduce the file size if a workbook has many cells that are duplicates. 499

  • Initial implementation of wb_add_pivot_table(). This allows adding native pivot tables to openxlsx2 workbooks. The pivot table area will remain empty until the sheet is opened in spreadsheet software and evaluated successfully. This feature is newly developed and can cause unexpected side effects. Be aware that using it might currently break workbooks.

Fixes

  • Reading of files with frozen panes and more than one section node was restored. 495

  • Fixed a copy and paste mistake in add_border() which used left borders for right borders. 496

  • Improve XML unescaping. 497

  • Fix reading and saving workbooks with multiple slicers per sheet. 505

  • Fix tab selection always selecting the first sheet since #303. 506

Breaking changes

  • Do not export write_data2() anymore. This was used in development in the early stages of the package and should not be used directly anymore.

  • Only documentation: openxlsx2 defaults to American English ‘color’ from now on. Though, we fully support the previous ‘colour’. Users will not have to adjust their code. Our documentation only lists color, but you can pass colour just the same way you used to. 501 502

openxlsx2 0.4.1

CRAN release: 2022-12-18

New features

  • Provide new argument calc_chain to wb_load(). This is set to FALSE per default, to ignore the calculation chain if it is found. This change only reflects files imported from third party spreadsheet software and should not be visible to the user. 461

  • Tweaks to wb_data(). Dims is now optional and will select data similar to wb_to_df(), similar it allows passing down other wb_to_df() arguments. Though, it probably is a good idea not be to creative passing down arguments, not all will result in a usable wb_data object. 462

  • Add hidden argument and change the default for heights to NULL in set_row_heights(). This allows changing the row height and/or hiding selected rows. This does not yet provide a way to hide rows per default. 475

  • Add wb_add_chartsheet() for chart sheet support. Along with internal cleanup around chart sheet code. 466

Fixes

  • Fix wb_freeze_pane(). This changes the load logic a bit. Previously we put everything into sheetViews (the frozen pane is part of this). Though wb_freeze_pane() assumes that freezePane is used. We now try to be smart and split sheetViews upon loading. 465

  • Previously, adding mschart objects to sheets was only possible if (1) the worksheet already contained a drawing (if the workbook was loaded) or (2) to the last sheet of the workbook. This has now been fixed. Adding mschart objects to any worksheet in the workbook is now possible as intended. 458

  • Really fix double xml escaping when saving. 467

  • Improving the drawing logic. There are some workbooks with various drawings per sheet and previously there were combinations possible that were not reflecting this. 478

openxlsx2 0.4

CRAN release: 2022-12-05

New features

  • Provide rvg support via wb_add_drawing(). This allows integrating rvg plots into xlsx files. 449

  • Improve print options. Defaults to printing grid lines, if the worksheet contains grid lines. 440

  • Support reading files with form control. 426

  • Handle input files with chart extensions. 443

  • Improve writing styles to workbook. Previously every cell was checked, this has been changed to check unique styles. 423

  • Implement reading custom file properties. 418

  • Improved add_named_region(). This function includes now various xml options. 386

  • Add … as argument to read_xlsx() and wb_read(). 381

  • Allow reading files with xml namespace created by third party software. 405

Fixes

  • Update or remove calculation chain when overwriting formulas in a workbook. 438

  • Fix double xml escaping when saving. 435

  • Minor tweak for POSIXct dates and try to avoid the notorious 29Feb1900. 424

  • Implement reading customXml folder for input files with connection. 419

  • Fixed saving files with <sheetPr/> tag. Previously this was wrapped in a second sheetPr node. This issue occurs with xlsm files only. 417

  • Fixed a case where embedded files were assigned incorrectly in worksheet relationships. This caused corrupted output. 403

Breaking changes

  • Remove merge_ functions for styles. 450

  • Previously if a loaded workbook contained formulas pointing to cells modified by openxlsx2, these formulas were not updated, once the workbook was opened in spreadsheet software. This is now enforced, unless the option openxlsx2.disableFullCalcOnLoad is set. In this case we would respect the original calculation properties of the workbook.

  • wb_save() no longer returns the path that the object was saved to, but instead the wbWorkbook object, invisibly. This is consistent with the behavior of others wrappers. 378

  • Remove never used all.equal.wbWorkbook(). The idea was nice, but it never developed into something useful.

  • Remove never used wb_chart_sheet() function. 399

Internal changes

  • Provide set_sheetview() in sheets. Can be used to provide a wbWorkbook function and wrapper in the future. 399

openxlsx2 0.3.1

CRAN release: 2022-10-31

New features

  • Functions adding data to a workbook now bring an applyCellStyle argument. If this is TRUE openxlsx2 will apply a numeric style, if FALSE we will simply add the numeric value without additional styling and use the previous cell style. 365

  • Reading from file or workbook with showFormula now returns all formulas found in the workbook. Previously only those with type e or str were returned. Now we will be able to see formulas like hyperlinks too. 352

Internal changes

  • Moved data validation list from x14 to data validation. This enables data validation lists without x14 extension openxlsx 386. 347

  • Removed build_cell_merges() and replaced it with a plain R solution. 390

Fixes

  • Improvements to setting column widths. Previously values set by set_col_widths() were a little off. This has now been improved. There are still corner cases where the column width set with openxlsx2 does not match those shown in spreadsheet software. Notable differences can be seen with floating point values (e.g., 10L works while 10.1 is slightly off) and with column width on Mac. 350

  • Improve rowNames when writing data to worksheet. Previously the name for the rownames column defaulted to 1. This has been changed. Now with data it defaults to an empty cell and with a data table it defaults to _rowNames_. 375

  • Fix the workbook xml relationship file to not include a reference to shared strings per default. This solves 360 for plain data files written from openxlsx2. 363

  • Adding cell styles has been streamlined to increase consistency. This includes all style functions like wb_add_font() and covers all cases of hyperlinks. 365

  • Fix cloning pivot charts. 361

  • Fix loading and writing files with slicers. Loading would add a few empty slicer xml files to Content_Types and workbook.xml.rels. 361

  • Align the logic for writing data to empty worksheets and updating/writing to worksheets with data. This removes update_cell_loop() and changes how update_cell() behaves. Not only does this remove duplicated code, it also brings great speed improvements (issue 356). 356

  • It is now possible to use special characters in formulas without coding. Previously & had to be encoded like &amp; 251

Breaking changes

  • Previously deprecated names.wbWorkbook() and names<-.wbWorkbook() have been removed. 367

  • Conditional style defaults for create_dxfs_style() have changed to be more permissive. Previously we shipped a default font, default font size and font color. This has been changed to better reflect a behavior the user expects. 343

openxlsx2 0.3

CRAN release: 2022-09-30

New features

  • New argument startCol in read to data frame functions wb_to_df(), wb_read() and read_xlsx(). 330

  • New function wb_colour() to ease working with color vectors used in openxlsx2 styles. 292

  • Deprecated get_cell_style() and set_cell_style() in favor of newly introduced wrapper functions wb_get_cell_style() and wb_set_cell_style(). 306

  • Improvements to wb_clone_worksheet(). Cloning of chartsheets as well as worksheets containing charts, pivot tables, drawings and tables is now possible or tweaked. Slicers are removed from the cloned worksheet. 305

  • Allow writing class data.table. 313

  • Provide na.numbers for reading functions, that convert numbers to NA in R output. Handle na.strings in write_xlsx(). 301

  • Add new option to add sparklines with various style options to worksheets: wb_add_sparklines(). Sparklines can be created with create_sparklines(). The manual page contains an example. 280

  • Add new options to data validation. allow type custom, add arguments errorStyle, errorTitle, error, promptTitle, prompt. 271

  • Provide wb_clone_sheet_style(). This improves upon the now deprecatedcloneSheetStyle() that existed as an early draft. 233

  • wb$add_data() now checks earlier for missing x argument. 246

Internal changes

  • Worksheets added to a wbWorkbook no longer contain default references to the drawings and vmlDrawings directories. Previously, these references were added as rId1 and rId2 even if the worksheet did not contain any drawing (e.g., an image or a chart) or vml drawing (e.g., a comment or a button). In such cases certain third party software, strictly following the references in worksheet or Content_Types complained about missing files and the import of such files failed completely. 311

  • Implement loading of user defined chartShapes. Previously this was not implemented instead the previous logic assumed that every sheet has a matching drawing. With chartShapes this no longer is true. The number of drawings and the number of worksheets/chartsheets must not match. 323

  • When loading files with charts, they are now imported into the wbWorkbook object. Previously they were simply copied. This will allow easier interaction with charts in the future. 304

  • Moving the data validation code from the workbook to the worksheet. Also, data_validation_list() is no longer stored in dataValidationLst. It has been moved to extLst, fixing a bug when saving and adding another data validation list. The code for retrieving the date origin from a workbook has been improved and get_date_origin(wb, origin = TRUE) now returns the origin as an integer from a wbWorkbook. 299

  • Removed level4 from XML functions. There was only a single use case for a level4 function that has been solved differently. If level4 is needed, this can be solved using a level3 and additional level2 functions. In addition xml_nodes now return nodes for all reachable nestings, therefore xml_node("<a/><a/>", "a") will now return a character vector of length two. For xml_node("<a/><b/>", "a") only a single character vector is returned. 280

  • Changes to various internal pugixml functions, to improve handling of XML strings. 279

  • Provide internal helper xml_rm_child() to remove children of XML strings. 273

  • Fixes a bug in update_cell() that slowed down writing on worksheets with data. In addition, this function has been cleaned up and further improved. It is no longer exported, as users only need to use wb_add_data() or write_data(), each of which calls update_cell() under the hood. 275 276

  • Various (mostly internal) changes to conditional_formatting. Created style_mgr integration for dxf (cf-styles) and cleaned up internal code. The syntax has changed slightly, see conditional formatting vignette for reference. Add whitespace argument to read_xml(). 268

Breaking changes

openxlsx2 0.2.1

New features

  • Data adding functions now ship a dims argument that can be used to determine the startCol and startRow for any x object that is added to the worksheet. Works with add_data(), add_data_table(), add_formula() and their underlying write_ functions as well as with the wrappers.

  • Provide optional na.strings argument when writing data to sheets. It can be used to add a custom character string when writing numeric data.

  • Improve writing NA, NaN, and -Inf/Inf. NA will be converted to #N/A; NaN will be converted to #VALUE!; Inf will be converted to #NUM!. The same conversion is not applied when reading from a workbook. 256

  • Many wbWorkbook methods now contain default sheet values of current_sheet() or next_sheet() (e.g., $add_worksheet(sheet = next_sheet()), $write_data(sheet = curret_sheet()). These internal waiver functions allow the wbWorkbook object to use default expectations for what sheet to interact with. This allows the easier workflow of wb$add_worksheet()$add_data(x = data.frame()) where $add_worksheet() knows to add a new worksheet (with a default name), sets that new worksheet to the current worksheet, and then $add_data() picks up the new sheet and places the data there. 165, 179

  • New functions wb_add_cell_style() and wb$add_cell_style() to simplify the creation of cell styles for cells on the sheet. This provides a fast way to create cell styles for regions on the worksheet. The cells for which the cell format is to be created must already exist on the worksheet. If the cells already contain a cell format, it will be preserved, except for the updated cell format entries, which will always be created. The function is applied to a continuous cell of the worksheet. 230

  • New functions wb_add_numfmt() and wb$add_numfmt() to simplify the creation of number formats for cells on the sheet. This provides a fast way to create number formats for regions on the worksheet. The cells for which the number format is to be created must already exist on the worksheet. If the cells already contain a cell style, it will be preserved, except for the number format, which will always be created. The function is applied to a continuous cell of the worksheet. 229

  • New functions wb_add_font() and wb$add_font() to simplify the creation of fonts for cells on the sheet. This provides a fast way to create fonts for regions on the worksheet. The cells for which the font is to be created must already exist on the worksheet. If the cells already contain a cell style, it will be preserved, except for the font, which will always be created. The function is applied to a continuous cell of the worksheet. 228

  • New functions wb_add_fill() and wb$add_fill() to simplify the creation of fills for cells on the sheet. This provides a fast way to create color filled regions on the worksheet. The cells for which the fill is to be created must already exist on the worksheet. If the cells already contain a cell style, it will be preserved, except for the filled color, which will always be created. The function is applied to a continuous cell of the worksheet and allows to change the color of every n-th column or row. 222

  • New functions wb_add_border() and wb$add_border() to simplify the creation of borders for cells on the sheet. This is especially useful when creating surrounding borders with different border styles for various cells. The cells for which the border is to be created must already exist on the worksheet. If the cells already contain a cell style, it will be preserved, except for the border, which will always be created. The function is applied to a continuous cell of the worksheet and allows to change the horizontal and vertical internal border grid independently. 220

  • Enable reading tables with wb_to_df(). Tables are handled similar to defined names. 193

  • Several enhancements have been added for checking and validation worksheet names 165

    • When adding a new worksheet via wbWorkbook$add_worksheet() the provided name is checked for illegal characters (see note in Breaking changes)
    • wbWorkbook$get_sheet_names() (wb_get_sheet_names() wrapper) added. These return both the formatted and original sheet names
    • wbWorkbook$set_sheet_names() (wb_set_sheet_names()) added
      • these make names.wbWorkbook() and names<-.wbWorkbook() deprecated
      • wbWorkbook$setSheetName() deprecated
    • clean_worksheet_names() added to support removing characters that are not allowed in worksheet names

Fixes

  • Various fixes to enable handling of non unicode R environments 243

  • Fix an issue with broken pageSetup reference causing corrupt excel files 216

  • Fix reading and writing comments from workbooks that already provide comments 209

  • Fix an issue with broken xml in Excels vml files and enable opening xlsm files with wb$open() 202

  • Fix reading and writing on non UTF-8 systems 198 199 207

  • Instruct parser to import nodes with whitespaces. This fixes a complaint in spreadsheet software. 189

  • Fix reading file without row attribute. 187 190

  • Remove reference to printerSettings.bin when loading. This binary blob is not included and the reference caused file corruption warnings. 185

  • Fix loading and writing xlsx files with with workbook$extLst. Previously if the loaded sheet contains a slicer, a second extLst was added which confused spreadsheet software. Now both are combined into a single node.

  • Fix writing xlsx file with multiple entries of conditional formatting type databar on any sheet. 174

  • Cell fields cm, ph and vm are now implemented for reading and writing. This is the first step to handle functions that use metadata. 173

  • wbWorkbook: $open() no longer overwrites the $path field to the temporary file 171

  • xl_open() works (better) on Windows 170

Breaking changes

  • When writing to existing workbooks, the default value for removeCellStyle is now FALSE. Therefore if a cell contains a style, it is attempted to replace the value, but not the style of the cell itself.

  • wb_conditional_formatting() is deprecated in favor of wb_add_conditional_formatting() and wbWorkbook$add_conditional_formatting().

    • type must now match exactly one of: "expression", "colorScale", "dataBar", "duplicatedValues", "containsText", "notContainsText", "beginsWith", "endsWith", "between", "topN", "bottomN"

  • Assigning a new worksheet with an illegal character now prompts an error 165. See ?clean_worksheet_name for an easy method of replacing bad characters.

  • openxlsx2Coerce() (which was called on x objects when adding data to a workbook) has been removed. Users can no longer pass some arbitrary objects and will need to format these objects appropriately or rely on as.data.frame methods 167

  • xl_open(file = ) is no longer valid and will throw a warning; first argument has been changes to x to highlight that xl_open() can be called on a file path or a wbWorkbook object 171

Internal changes

  • Remove wb$createFontNode() which was never used.

  • Switch to modern xlsx template, when creating workbooks. Imported workbooks will use the imported template

  • Rewrite wb$tables to use a data frame approach. This simplifies the code a bit and makes it easier to implement more upcoming changes 191

  • Update of internal pugixml library

  • The two functions write_data() and write_datatable() now use the same internal function write_data_table() to add data to the sheet. This simplifies the code and ensures that both functions are tested. In the same pull request, the documentation has been updated and the stack= option, which was not present before, has been removed 175

  • wbWorkbook$validate_sheet() added as an object methods

  • private wbWorkbook field original_sheet_names added to track the original names passed to sheets

  • private $get_sheet() removed in favor of more explicit

  • private wbWorkbook methods additions:

    • $get_sheet_id_max(), $get_sheet_index() for getting ids
    • $get_sheet_name() for getting a sheet name
    • $set_single_sheet_name() for setting sheet names
    • $pappend() general private appending
    • $validate_new_sheet() for checking new sheet names
    • $append_workbook_field() for self$workbook[[field]]
    • $append_sheet_rels() for self$worksheet_rels[[sheet]]
    • $get_worksheet() to replace $ws()

openxlsx2 0.2.0

  • Added a NEWS.md file to track changes to the package.
  • First public release