Unverified Commit 89b9a50d authored by John Davis's avatar John Davis Committed by GitHub
Browse files

Merge pull request #20827 from arash77/add-pptx-datatype 

[25.0] Fix DOCX detection and add PPTX support
parents c47bb0c4 98fae115

lib/galaxy/config/sample/datatypes_conf.xml.sample

+5 −0
Original line number Diff line number Diff line
@@ -695,6 +695,9 @@ 
    <datatype extension="docx" type="galaxy.datatypes.binary:Docx" display_in_upload="true" decription="DOCX is an XML-based file format that is natively used for Microsoft Word documents" description_url="https://www.iso.org/standard/71691.html">

        <infer_from suffix="docx" />

    </datatype>

    <datatype extension="pptx" type="galaxy.datatypes.binary:Pptx" display_in_upload="true" decription="PPTX is an XML-based file format that is natively used for Microsoft PowerPoint presentations" description_url="https://www.iso.org/standard/71691.html">

        <infer_from suffix="pptx" />

    </datatype>

    <datatype extension="btwisted" type="galaxy.datatypes.data:Text" subclass="true"/>

    <datatype extension="cai" type="galaxy.datatypes.data:Text" subclass="true"/>

    <datatype extension="cat_db" type="galaxy.datatypes.data:Text" subclass="true"/>
@@ -1270,6 +1273,8 @@ 
    <sniffer type="galaxy.datatypes.binary:Edr"/>

    <sniffer type="galaxy.datatypes.binary:Vel"/>

    <sniffer type="galaxy.datatypes.binary:Xlsx"/>

    <sniffer type="galaxy.datatypes.binary:Docx"/>

    <sniffer type="galaxy.datatypes.binary:Pptx"/>

    <sniffer type="galaxy.datatypes.binary:Numpy"/>

    <sniffer type="galaxy.datatypes.qiime2:QIIME2Metadata"/>

    <sniffer type="galaxy.datatypes.qiime2:QIIME2Artifact"/>

lib/galaxy/datatypes/binary.py

+17 −0
Original line number Diff line number Diff line
@@ -3184,6 +3184,7 @@ class Docx(Binary): 


    file_ext = "docx"

    compressed = True

    display_behavior = "download"  # Office documents trigger downloads



    def sniff_prefix(self, file_prefix: FilePrefix) -> bool:

        # Docx is compressed in zip format and must not be uncompressed in Galaxy.
@@ -3206,6 +3207,22 @@ class Xlsx(Binary): 
        return file_prefix.compressed_mime_type == "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"





@build_sniff_from_prefix

class Pptx(Binary):

    """Class for PowerPoint 2007 (pptx) files"""



    file_ext = "pptx"

    compressed = True

    display_behavior = "download"  # Office documents trigger downloads



    def sniff_prefix(self, file_prefix: FilePrefix) -> bool:

        # Pptx is compressed in zip format and must not be uncompressed in Galaxy.

        return (

            file_prefix.compressed_mime_type

            == "application/vnd.openxmlformats-officedocument.presentationml.presentation"

        )





@build_sniff_from_prefix

class ExcelXls(Binary):

    """Class describing an Excel (xls) file"""

lib/galaxy/datatypes/sniff.py

+6 −1
Original line number Diff line number Diff line
@@ -52,7 +52,12 @@ if TYPE_CHECKING: 
log = logging.getLogger(__name__)



SNIFF_PREFIX_BYTES = int(os.environ.get("GALAXY_SNIFF_PREFIX_BYTES", None) or 2**20)

BINARY_MIMETYPES = {"application/pdf", "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"}

BINARY_MIMETYPES = {

    "application/pdf",

    "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",

    "application/vnd.openxmlformats-officedocument.wordprocessingml.document",

    "application/vnd.openxmlformats-officedocument.presentationml.presentation",

}





def get_test_fname(fname):