Paragraph Spacing - lineRule options

[DKWoods]

I recently discovered some DOCx files that specify the paragraph spacing a bit differently. Instead of <w:spacing w:line="240" w:lineRule="exact"/>, it uses <w:spacing w:line="240" w:lineRule="exactly"/>, using a different value for the lineRule parameter.

My copy of Word 2010 uses 'lineRule="exact"', as does python-docx. I have not been able to find any documentation online that specifies allowable values for this parameter. Does anyone know where I can find out if "exactly" is a synonym, perhaps used by a later version of Word? The only other possibility I can think of is that the program producing these DOCx files is doing it wrong, but that's not really something I have control over.

[scanny]

This would be the realm of the ISO-29500 spec, and probably most usefully in the XSD (Xml Schema) files that accompany it. A copy of those are in the code tree here:
https://github.com/python-openxml/python-docx/tree/master/ref/xsd

Most prose information is in the ISO-IEC-29500-1.pdf file:
https://github.com/python-openxml/python-docx/blob/master/ref/ISO-IEC-29500-1.pdf

The xsd directory contains the Xml Schema files. wml.xsd is the core WordprocessingML schema:
https://github.com/python-openxml/python-docx/blob/master/ref/xsd/wml.xsd

The definition related to w:lineRule is ST_LineSpacingRule which is here: https://github.com/python-openxml/python-docx/blob/master/ref/xsd/wml.xsd#L601

... and "exactly" is not one of the options.

So either the spec has changed (doubtful in my mind) or the documents are being generated by something with a bit of a bug in it. I suppose it's possible that the other spec (ECMA 376) differs on this count, you'll need to Googlize all of that.

Does Word do the right thing (like space them correctly) with files that come in that way?

[DKWoods]

Thanks Steve.  This is really helpful. ISO-IEC-295000-1.pdf, in the ref directory, lists the value as "exactly" on page 252 at the bottom.  wml.xsd clearly uses "exact", as you have said.  ECMA 376 agrees with xml.xsd that the value is "exact", not "exactly". When I create a simple example document in Word 2010 and replace "exact" with "exactly" in document.xml, then load the document, the line spacing in the Paragraph Formatting dialog is displayed as "Multiple" rather than "Exactly".  So there's that. Behaviorally, at least for Word 2010, the parameter must be "exact". The source documents which are causing me this problem turn out to be converted PDF files which mix images and text.  The lineRule specs seem to appear in empty paragraphs preceding image specifications, so I'm guessing they're used in an attempt to mimic the PDF's layout.  They are honestly kind of a mess, as I look at them more closely. So the other guy is wrong.  That and a dollar will buy me ... what, a quarter cup of coffee maybe?  I still have to support import of his documents, and based on previous experience, the developer of the other product is ... not open to suggestions for improving his code at this level.  It may well not even be his code that is introducing this issue. So on a practical level, I suppose I will have to trap the error and ignore the incorrect value.  Right now, it raises an exception and blocks import of the document entirely.  Or maybe I should try to add "exactly" as a synonym for "exact" for lineRule. Thanks for the insight.  I feel better knowing the correct answer, even if I'm not in a position to force the issue. David

…

On 1/14/19 2:24 PM, Steve Canny wrote: This would be the realm of the ISO-29500 spec, and probably most usefully in the XSD (Xml Schema) files that accompany it. A copy of those are in the code tree here: https://github.com/python-openxml/python-docx/tree/master/ref/xsd Most prose information is in the ISO-IEC-29500-1.pdf file: https://github.com/python-openxml/python-docx/blob/master/ref/ISO-IEC-29500-1.pdf The |xsd| directory contains the Xml Schema files. |wml.xsd| is the core WordprocessingML schema: https://github.com/python-openxml/python-docx/blob/master/ref/xsd/wml.xsd The definition related to |w:lineRule| is |ST_LineSpacingRule| which is here: https://github.com/python-openxml/python-docx/blob/master/ref/xsd/wml.xsd#L601 ... and "exactly" is not one of the options. So either the spec has changed (doubtful in my mind) or the documents are being generated by something with a bit of a bug in it. I suppose it's possible that the other spec (ECMA 376) differs on this count, you'll need to Googlize all of that. Does Word do the right thing (like space them correctly) with files that come in that way? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#598 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AL5BDTC_IFmgWY3Uj2MsP0ew4o7uaZ-Dks5vDOeXgaJpZM4Z_mfO>.

[scanny]

Hmm. The discrepancy in the spec is a plausible explanation for how the error got there. Although the Xml Schema is the final word, not the spec. And the fact that Word doesn't accept it seals the case for me.

One thing you might consider is writing a transform on the input XML, or even just doing a search and replace as a pre-processing step. You could use opc-diag to extract the package into a directory and then just use command-line tools like sed or awk or whatever to do an edit on document.xml or whatever part, and then load the .docx straight from the directory (no need to repackage). I suppose all those steps could be bundled up into a shell script.

A transform (like an XSLT transform) would be more powerful, but much more sophisticated and perhaps not worth the trouble unless there were multiple other repairs involved.