Generated .docx get wrong DocProperty Field

[epernlu]

I try to generate a .docx file from markdown using a reference template but get issues with the Property of the Field "DocProperty" in my header. The Property is not the same in the result as in the template. Example files can be found here:

https://groups.google.com/forum/?authuser=0#!topic/pandoc-discuss/EqzYV7yO3Ss

If you for example look at "Checked" field in the header, the Property = Checked in the Template and = Bytes in the result file.

When I try to change Property in the generated document I get: Error! Unknown document property name.

Any suggestion on how this can be solved?

This is the command i run:

pandoc --reference-docx=Template.docx markdown.md -s -o result.doc

[nkalvi]

It looks like Word stores properties in docProps/custom.xml (inside the docx) when any of them are modified (docx without any changes to properties doesn't seem to contain custom.xml). Current version of Pandoc does not seem to be copying the following items (along with the corresponding references) to the resulting docx:

Template
├── customXml
│   ├── _rels
│   │   └── item1.xml.rels
│   ├── item1.xml
│   └── itemProps1.xml
├── docProps
│   └── custom.xml

Partial contents of the custom.xml

<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
<Properties
    xmlns="http://schemas.openxmlformats.org/officeDocument/2006/custom-properties"
    xmlns:vt="http://schemas.openxmlformats.org/officeDocument/2006/docPropsVTypes">
    <property fmtid="{D5CDD505-2E9C-101B-9397-08002B2CF9AE}" pid="2" name="SecurityClass">
        <vt:lpwstr></vt:lpwstr>
    </property>
    <property fmtid="{D5CDD505-2E9C-101B-9397-08002B2CF9AE}" pid="3" name="DocName">
        <vt:lpwstr></vt:lpwstr>
    </property>
    <property fmtid="{D5CDD505-2E9C-101B-9397-08002B2CF9AE}" pid="4" name="Prepared">
        <vt:lpwstr></vt:lpwstr>
    </property>
    <property fmtid="{D5CDD505-2E9C-101B-9397-08002B2CF9AE}" pid="5" name="DocNo">
        <vt:lpwstr></vt:lpwstr>
    </property>
    <property fmtid="{D5CDD505-2E9C-101B-9397-08002B2CF9AE}" pid="6" name="Revision">
        <vt:lpwstr></vt:lpwstr>
    </property>
    <property fmtid="{D5CDD505-2E9C-101B-9397-08002B2CF9AE}" pid="7" name="Checked">
        <vt:lpwstr></vt:lpwstr>

@jkr or @lierdakil may be able to explain or correct this.

In the mean time, if you're comfortable with zipping/unzipping docx (and have time), I could look into a workaround. Please let us now of you platform/Word version etc.

[nkalvi]

By the way, document fragment containing the example seems to be same in template.docx and result.docx (i.e. I don't think Pandoc is not altering the content in the document portion, so properly including the items above may correct the issue).

Template.docx

            <w:tc>
                <w:tcPr>
                    <w:tcW w:w="1318" w:type="dxa"/>
                    <w:tcBorders>
                        <w:left w:val="nil"/>
                        <w:bottom w:val="single" w:sz="6" w:space="0" w:color="auto"/>
                        <w:right w:val="single" w:sz="6" w:space="0" w:color="auto"/>
                    </w:tcBorders>
                </w:tcPr>
                <w:p w:rsidR="005326BC" w:rsidRPr="00EF6028" w:rsidRDefault="00270DA4">
                    <w:pPr>
                        <w:pStyle w:val="Header"/>
                    </w:pPr>
                    <w:r w:rsidRPr="00EF6028">
                        <w:fldChar w:fldCharType="begin"/>
                    </w:r>
                    <w:r w:rsidRPr="00EF6028">
                        <w:instrText xml:space="preserve"> DOCPROPERTY "Checked" \* MERGEFORMAT </w:instrText>
                    </w:r>
                    <w:r w:rsidRPr="00EF6028">
                        <w:fldChar w:fldCharType="end"/>
                    </w:r>
                </w:p>
            </w:tc>

Result.docx

            <w:tc>
                <w:tcPr>
                    <w:tcW w:w="1318" w:type="dxa"/>
                    <w:tcBorders>
                        <w:left w:val="nil"/>
                        <w:bottom w:val="single" w:sz="6" w:space="0" w:color="auto"/>
                        <w:right w:val="single" w:sz="6" w:space="0" w:color="auto"/>
                    </w:tcBorders>
                </w:tcPr>
                <w:p w:rsidR="005326BC" w:rsidRPr="00EF6028" w:rsidRDefault="00270DA4">
                    <w:pPr>
                        <w:pStyle w:val="Header"/>
                    </w:pPr>
                    <w:r w:rsidRPr="00EF6028">
                        <w:fldChar w:fldCharType="begin"/>
                    </w:r>
                    <w:r w:rsidRPr="00EF6028">
                        <w:instrText xml:space="preserve"> DOCPROPERTY "Checked" \* MERGEFORMAT </w:instrText>
                    </w:r>
                    <w:r w:rsidRPr="00EF6028">
                        <w:fldChar w:fldCharType="end"/>
                    </w:r>
                </w:p>
            </w:tc>

[epernlu]

Thanks for looking into this!! My platform is Windows 7 Enterprise and I am using MS Word 2010.

Did not know you could unzip a .docx file... =) Looking at that now!

[nkalvi]

You're most welcome @epernlu; I find this viewer very helpful in troubleshooting docx https://chrome.google.com/webstore/detail/ooxml-tools/bjmmjfdegplhkefakjkccocjanekbapn?hl=en.

[jgm]

Pandoc doesn't copy over this part of the reference.docx. We're pretty strict about what we bring over (and hence about what can be customized), because an earlier more relaxed policy frequently led to corrupt files. It might be worth experimenting with bringing over this custom stuff, to see if it can be done safely.

[agusmba]

Partially related to #3034 although with a different approach, here the property definition (and value) is in the template whereas in the linked issue, the property definition (and value) is in the yaml frontmatter, which seems to be a better general solution.