[Conversion Input] Doc Input Libreoffice Plugin

[sac]

Calibre plugin that makes docs searchable via Calibre full text search.

Description

The plugin enables Calibre internally to read and process doc files. The original doc in the user library remains unchanged. This is a fork of igieon/calibre-doc-input Thread.

Usage

If you're soffice.exe is not located in "C:\Program Files\LibreOffice\program\soffice.exe", you can edit it in HERE, then download the src folder, zip it (files at top level) & add it to Calibre.

IMPORTANT

Path to LibreOffice should be usually changed via Calibre Settings > Input Options > Doc Input > click on the location search button. This throws the following error:

File "calibre_plugins.doc_input.doc_input_ui", line 70, in fileSearch
TypeError: EnumType.__call__() missing 1 required positional argument: 'value'

If anyone has an idea what this is about, please let me know (here / fork / PR / etc.). I'm not really into python and just hacked this adjustment together in vscode based on igieon's great base.

[sac]

Github repository: calibre-doc-input-libreoffice

[sac]

Error seems fixed. Don't know how to edit the first post. Update is here.

[Quoth]

Better to simply open a .doc in LO Writer (save & edit as odt) or MS Word 2007 or later (save & edit in .docx) and fix page styles, paragraph styles, character styles, graphics, headers and links / anchors. Then a final extra save in docx for Calibre.

[sac]

Originally Posted by Quoth

Better to simply open a .doc in LO Writer (save & edit as odt) or MS Word 2007 or later (save & edit in .docx) and fix page styles, paragraph styles, character styles, graphics, headers and links / anchors. Then a final extra save in docx for Calibre.

Not an option for 200.000 docs. Also every conversion decreases quality. Example the attached original doc. LO can create a perfect 1:1 copy to ODT, but by converting to DOCX the headline is not intended and a wrong "headline (user)" instead of "headline" is used. Calibre itself is afterwards "removing" the first image of the docx if imported.

BTW: error is fixed, filedialog is showing, but 'wordconv_exe_path' is somehow still not used. It's always using the recommended value Probably because UI form is used instead of Widget?

[Quoth]

I've never seen a "doc" for windows file since 1993 that didn't need fixing, and now especially to make into an epub.

A perfect 1:1 is useless if the doc isn't styled, Headers and links/anchors properly done. Besides it doesn't do perfect conversions. It's not possible, MS Word 2007 and later don't either.

Same applies to most docx I'm given. People that don't know how to use Word.

[BetterRed]

Assuming you have Word 2007 or later you can bulk convert .doc files to .docx with a 22 line VBA macro ==>>> How to batch convert .doc files to .docx format in Word?

BR

[sac]

Originally Posted by BetterRed

Assuming you have Word 2007 or later you can bulk convert .doc files to .docx with a 22 line VBA macro ==>>> How to batch convert .doc files to .docx format in Word?

BR

That looks awesome, good to know

However, I have some archives with docs that I want to keep in their original form. Just need full-text search & administration in Calibre (and possibility to convert some in Calibre if required).

I rewrote the plugin as "class PluginWidget(Widget)" similar to KFX Input (that uses the same OptionRecommendation). However, issue remains that QT is not synching "options.wordconv_exe_path". Noone who has an idea what's causing this?

[BetterRed]

I do recall that the location of wordconv.exe was the bane of the DOC conversion input plugin.

The macro in the link I gave you creates a new .docx alongside the .doc. I used a similar macro (it had the folder path hard coded) to convert several thousand .doc files to .docx back in 2009. I deleted the .doc files soon after moving the .docx files into a calibre library a couple of years later.

BR

[sac]

That's why this plugin is using LibreOffice. Attached the Widget version, GUI shows now also on a fresh Calibre Portable in Settings > Input Options, but still:

issue remains that QT is not synching "options.wordconv_exe_path". Noone who has an idea what's causing this?

[sac]

In general it seems that the missing options argument in file dialogs was a bug in PyQt6, fixed in QtPy 2.4.0 https://github.com/spyder-ide/qtpy/releases/tag/v2.4.0 in 2023

I wonder what version Calibre uses.

[sac]

Calibre is not using QtPy.

Guess the error is somewhere in

__init__.py Imports or config

    def gui_configuration_widget(self, parent, get_option_by_name, get_option_help, db, book_id=None):
        from calibre_plugins.doc_input.doc_input import PluginWidget
        return PluginWidget(parent, get_option_by_name, get_option_help, db, book_id)

doc_input.py Imports or config

    def __init__(self, parent, get_option, get_help, db=None, book_id=None):
        self.db = db                # db is set for conversion, but not default preferences
        self.book_id = book_id      # book_id is set for individual conversion, but not bulk

        Widget.__init__(self, parent, ['docx_no_cover', 'wordconv_exe_path'])
        self.initialize_options(get_option, get_help, db, book_id)

Would be awesome if someone can confirm if this is correct.