From cacc48beb90706ce05805be98539a03a1cf66c54 Mon Sep 17 00:00:00 2001 From: lopez Date: Tue, 22 Oct 2019 00:12:22 +0200 Subject: [PATCH] remove doc --- .../java/org/grobid/core/engines/FullTextParser.java | 2 +- .../java/org/grobid/core/engines/HeaderParser.java | 4 ++-- .../grobid/service/process/GrobidRestProcessFiles.java | 10 +--------- 3 files changed, 4 insertions(+), 12 deletions(-) diff --git a/grobid-core/src/main/java/org/grobid/core/engines/FullTextParser.java b/grobid-core/src/main/java/org/grobid/core/engines/FullTextParser.java index 8d0e21aba4..e9c3521c6b 100755 --- a/grobid-core/src/main/java/org/grobid/core/engines/FullTextParser.java +++ b/grobid-core/src/main/java/org/grobid/core/engines/FullTextParser.java @@ -110,7 +110,7 @@ public Document processing(File input, GrobidAnalysisConfig config) throws Exception { DocumentSource documentSource = null; String extension = FilenameUtils.getExtension(input.getName()); - if ( extension != null && (extension.toLowerCase().equals("docx") || extension.toLowerCase().equals("doc")) ) { + if ( extension != null && (extension.toLowerCase().equals("docx")) ) { documentSource = DocumentSource.fromDocx(input, config.getStartPage(), config.getEndPage(), config.getPdfAssetPath() != null, true, false); } else { diff --git a/grobid-core/src/main/java/org/grobid/core/engines/HeaderParser.java b/grobid-core/src/main/java/org/grobid/core/engines/HeaderParser.java index 51501a30da..2619018ded 100755 --- a/grobid-core/src/main/java/org/grobid/core/engines/HeaderParser.java +++ b/grobid-core/src/main/java/org/grobid/core/engines/HeaderParser.java @@ -79,7 +79,7 @@ public Pair processing(File input, BiblioItem resHeader, Grobi DocumentSource documentSource = null; try { String extension = FilenameUtils.getExtension(input.getName()); - if ( extension != null && (extension.toLowerCase().equals("docx") || extension.toLowerCase().equals("doc")) ) { + if ( extension != null && (extension.toLowerCase().equals("docx")) ) { documentSource = DocumentSource.fromDocx(input, config.getStartPage(), config.getEndPage()); } else { documentSource = DocumentSource.fromPdf(input, config.getStartPage(), config.getEndPage()); @@ -104,7 +104,7 @@ public Pair processing2(String input, BiblioItem resHeader, Gr DocumentSource documentSource = null; try { String extension = FilenameUtils.getExtension(input); - if ( extension != null && (extension.toLowerCase().equals("docx") || extension.toLowerCase().equals("doc")) ) { + if ( extension != null && (extension.toLowerCase().equals("docx")) ) { documentSource = DocumentSource.fromDocx(new File(input), config.getStartPage(), config.getEndPage()); } else { diff --git a/grobid-service/src/main/java/org/grobid/service/process/GrobidRestProcessFiles.java b/grobid-service/src/main/java/org/grobid/service/process/GrobidRestProcessFiles.java index b1c157b77f..943644d6d3 100644 --- a/grobid-service/src/main/java/org/grobid/service/process/GrobidRestProcessFiles.java +++ b/grobid-service/src/main/java/org/grobid/service/process/GrobidRestProcessFiles.java @@ -86,10 +86,8 @@ public Response processStatelessHeaderDocument(final InputStream inputStream, fi String fileNameLow = fileName.toLowerCase(); if (fileNameLow.endsWith("docx")) extension = "docx"; - else if (fileNameLow.endsWith("doc")) - extension = "doc"; } - + // the tmp file will have an unambiguous file extension (e.g. .pdf. .docx, .doc, ...) originFile = IOUtilities.writeInputFile(inputStream, extension); if (originFile == null) { @@ -179,8 +177,6 @@ public Response processFulltextDocument(final InputStream inputStream, String fileNameLow = fileName.toLowerCase(); if (fileNameLow.endsWith("docx")) extension = "docx"; - else if (fileNameLow.endsWith("doc")) - extension = "doc"; } // the tmp file will have an unambiguous file extension (e.g. .pdf. .docx, .doc, ...) @@ -280,8 +276,6 @@ public Response processStatelessFulltextAssetDocument(final InputStream inputStr String fileNameLow = fileName.toLowerCase(); if (fileNameLow.endsWith("docx")) extension = "docx"; - else if (fileNameLow.endsWith("doc")) - extension = "doc"; } // the tmp file will have an unambiguous file extension (e.g. .pdf. .docx, .doc, ...) @@ -540,8 +534,6 @@ public Response processStatelessReferencesDocument(final InputStream inputStream String fileNameLow = fileName.toLowerCase(); if (fileNameLow.endsWith("docx")) extension = "docx"; - else if (fileNameLow.endsWith("doc")) - extension = "doc"; } originFile = IOUtilities.writeInputFile(inputStream, extension);