[refactor] Confirm parts are correct when citations have duplicate fi…

…le names (#655) Co-authored-by: Ian Seabock (Centific Technologies Inc) <[email protected]>
microsoft · Feb 27, 2024 · 16c260c · 16c260c
1 parent 5679b31
commit 16c260c
Show file tree

Hide file tree

Showing 6 changed files with 26 additions and 9 deletions.
diff --git a/frontend/src/api/models.ts b/frontend/src/api/models.ts
@@ -7,6 +7,7 @@ export type AskResponse = {
 };
 
 export type Citation = {
+    part_index?: number;
     content: string;
     id: string;
     title: string | null;

diff --git a/frontend/src/components/Answer/Answer.tsx b/frontend/src/components/Answer/Answer.tsx
@@ -68,13 +68,14 @@ export const Answer = ({
     const createCitationFilepath = (citation: Citation, index: number, truncate: boolean = false) => {
         let citationFilename = "";
 
-        if (citation.filepath && citation.chunk_id) {
+        if (citation.filepath) {
+            const part_i = citation.part_index ?? (citation.chunk_id ? parseInt(citation.chunk_id) + 1 : '');
             if (truncate && citation.filepath.length > filePathTruncationLimit) {
                 const citationLength = citation.filepath.length;
-                citationFilename = `${citation.filepath.substring(0, 20)}...${citation.filepath.substring(citationLength -20)} - Part ${parseInt(citation.chunk_id) + 1}`;
+                citationFilename = `${citation.filepath.substring(0, 20)}...${citation.filepath.substring(citationLength - 20)} - Part ${part_i}`;
             }
             else {
-                citationFilename = `${citation.filepath} - Part ${parseInt(citation.chunk_id) + 1}`;
+                citationFilename = `${citation.filepath} - Part ${part_i}`;
             }
         }
         else if (citation.filepath && citation.reindex_id) {

diff --git a/frontend/src/components/Answer/AnswerParser.tsx b/frontend/src/components/Answer/AnswerParser.tsx
@@ -7,6 +7,20 @@ type ParsedAnswer = {
     markdownFormatText: string;
 };
 
+const enumerateCitations = (citations: Citation[]) => {
+    const filepathMap = new Map();
+    for (const citation of citations) {
+        const { filepath } = citation;
+        let part_i = 1
+        if (filepathMap.has(filepath)) {
+            part_i = filepathMap.get(filepath) + 1;
+        }
+        filepathMap.set(filepath, part_i);
+        citation.part_index = part_i;
+    }
+    return citations;
+}
+
 export function parseAnswer(answer: AskResponse): ParsedAnswer {
     let answerText = answer.answer;
     const citationLinks = answerText.match(/\[(doc\d\d?\d?)]/g);
@@ -27,6 +41,7 @@ export function parseAnswer(answer: AskResponse): ParsedAnswer {
         }
     })
 
+    filteredCitations = enumerateCitations(filteredCitations);
 
     return {
         citations: filteredCitations,

diff --git a/static/assets/index-b152f819.js → static/assets/index-cb20ac21.js b/static/assets/index-b152f819.js → static/assets/index-cb20ac21.js
diff --git a/static/assets/index-b152f819.js.map → static/assets/index-cb20ac21.js.map b/static/assets/index-b152f819.js.map → static/assets/index-cb20ac21.js.map
diff --git a/static/index.html b/static/index.html
@@ -5,7 +5,7 @@
     <link rel="icon" type="image/x-icon" href="{{ favicon }}" />
     <meta name="viewport" content="width=device-width, initial-scale=1.0" />
     <title>{{ title }}</title>
-    <script type="module" crossorigin src="/assets/index-b152f819.js"></script>
+    <script type="module" crossorigin src="/assets/index-cb20ac21.js"></script>
     <link rel="stylesheet" href="/assets/index-0774d4f6.css">
   </head>
   <body>