Merge pull request #36949 from overleaf/em-bibtex-crlf-upload-fix

[web] fix bibtex parser dropping fields on CRLF/CR line endings

GitOrigin-RevId: 6174bedeed92a014c2d7210afb6f39387e8608a8

services/web/frontend/js/features/bibtex/bibtex-parser.ts

@@ -75,7 +75,7 @@ export class BibtexEntryAccumulator {
     if (type.is(StringLiteral)) {
       // Strip surrounding quotes/braces and collapse multi-line indentation.
       let s = getText(node.from + 1, node.to - 1)
-      s = s.replaceAll(/\n\s*/g, ' ')
+      s = s.replaceAll(/[\n\r]\s*/g, ' ')
       this.fieldValue = this.fieldValue.addString(s)
       return false
     }

services/web/frontend/js/features/source-editor/lezer-bibtex/bibtex.grammar

@@ -1,12 +1,12 @@
 @tokens {
-  whiteSpace { $[\t\n ]+ }
+  whiteSpace { $[\t\n\r ]+ }
 
   // Anything outside of an entry at the top level is a BibTeX free-form comment
-  Junk { ![@ \t\n%] ![@]* }
+  Junk { ![@ \t\n\r%] ![@]* }
 
   // BibLaTeX supports comments starting with a % and extending to the end of
   // the line
-  Comment { "%" ![\n]* }
+  Comment { "%" ![\n\r]* }
 
   // See https://metacpan.org/release/AMBS/Text-BibTeX-0.91/view/btparse/doc/bt_language.pod
   identifierSymbol { $[!$&*+./:;<>?^`_|] | "[" | "]" | "-" }

services/web/frontend/js/ide/log-parser/bib-log-parser.ts

@@ -85,7 +85,7 @@ export default class BibLogParser {
     }
     this.text = text.replace(/(\r\n)|\r/g, '\n')
     this.options = options
-    this.lines = text.split('\n')
+    this.lines = this.text.split('\n')
 
     // each parser is a pair of [regex, processFunction], where processFunction
     // describes how to transform the regex mactch into a log entry object.

services/web/test/frontend/ide/log-parser/logParserTests.js

@@ -296,6 +296,15 @@ describe('logParser', function () {
     )
   })
 
+  it('should parse a biber .blg file with CRLF line endings', function () {
+    const log = readLog('biber.blg').replace(/\n/g, '\r\n')
+    const bibParser = new BibLogParser(log, {})
+    const result = bibParser.parse()
+    expect(result.all.length).to.equal(14)
+    expect(result.errors.length).to.equal(1)
+    expect(result.warnings.length).to.equal(2)
+  })
+
   it('should throw an error when non-biblog passed to BibLogParser', function () {
     const log = readLog('caption-warnings.log')
     const bibParser = new BibLogParser(log, {})