commit dff586735b6618d9b011823594a33287d8f7f223
author Mike Kaganski <mike.kaganski@collabora.com> Mon May 03 17:04:04 2021 +0200
committer Mike Kaganski <mike.kaganski@collabora.com> Mon May 03 18:38:38 2021 +0200
tree b68025662bc28e5fbb9c0af689b74e719e86a68a
parent 9f89ee7c5076f700589d3b07f3d6a50f9af7d13a

tdf#123476: also use filter by extension when its service is the same

... as passed explicitly in the media descriptor, for 0-byte files.
This makes proper handling for such files when passing module name
in command line, like --writer for DOCX.

Change-Id: If8fd51e65dcf8a67b2653026f5fc1d5c964074af
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114924
Tested-by: Jenkins
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>

filter/source/textfilterdetect/filterdetect.cxx

diff --git a/filter/source/textfilterdetect/filterdetect.cxx b/filter/source/textfilterdetect/filterdetect.cxx
index 09efe5e5..19696ec 100644
--- a/filter/source/textfilterdetect/filterdetect.cxx
+++ b/filter/source/textfilterdetect/filterdetect.cxx
@@ -135,7 +135,7 @@ bool IsHTMLStream( const uno::Reference<io::XInputStream>& xInStream )
 * writes the type name to rType, the filter name to rMediaDesc.
 */
bool HandleEmptyFileUrlByExtension(MediaDescriptor& rMediaDesc, const OUString& rExt,
                                   OUString& rType)
                                   OUString& rType, OUString& rService)
{
    OUString aURL = rMediaDesc.getUnpackedValueOrDefault(MediaDescriptor::PROP_URL(), OUString());
    if (!tools::isEmptyFileUrl(aURL))
@@ -156,6 +156,7 @@ bool HandleEmptyFileUrlByExtension(MediaDescriptor& rMediaDesc, const OUString& 

    rMediaDesc[MediaDescriptor::PROP_FILTERNAME()] <<= pFilter->GetFilterName();
    rType = pFilter->GetTypeName();
    rService = pFilter->GetServiceName();
    return true;
}
}
@@ -216,14 +217,25 @@ OUString SAL_CALL PlainTextFilterDetect::detect(uno::Sequence<beans::PropertyVal
        OUString aName = aParser.getName().toAsciiLowerCase();

        // Decide which filter to use based on the document service first,
        // then on extension if that's not available.
        // then on extension if that's not available. Make exception for 0-byte files
        // whose extensions are handled by the same service as passed document service.
        OUString aEmptyType, aEmptyService;
        bool bEmpty = HandleEmptyFileUrlByExtension(aMediaDesc, aExt, aEmptyType, aEmptyService);
        if (bEmpty && aDocService == aEmptyService)
        {
            aDocService.clear(); // don't fallback to text filter, use extension-based match
            // TODO: maybe reset aExt when it's "xls"
        }

        if (aDocService == CALC_DOCSERVICE)
            aMediaDesc[MediaDescriptor::PROP_FILTERNAME()] <<= OUString(CALC_TEXT_FILTER);
        else if (aDocService == WRITER_DOCSERVICE)
            aMediaDesc[MediaDescriptor::PROP_FILTERNAME()] <<= OUString(WRITER_TEXT_FILTER);
        else if (aExt == "csv" || aExt == "tsv" || aExt == "tab" || aExt == "xls" || aName.endsWith(".csv.gz"))
            aMediaDesc[MediaDescriptor::PROP_FILTERNAME()] <<= OUString(CALC_TEXT_FILTER);
        else if (!HandleEmptyFileUrlByExtension(aMediaDesc, aExt, aType))
        else if (bEmpty)
            aType = aEmptyType; // aMediaDesc is already updated in HandleEmptyFileUrlByExtension
        else
            aMediaDesc[MediaDescriptor::PROP_FILTERNAME()] <<= OUString(WRITER_TEXT_FILTER);
    }