pi-ocr 1.3.0 → 1.3.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/extensions/mineru-pro.ts +10 -3
- package/package.json +1 -1
package/extensions/mineru-pro.ts
CHANGED
|
@@ -144,7 +144,7 @@ async function pollSingleTask(
|
|
|
144
144
|
const state: string = data.state || "unknown";
|
|
145
145
|
|
|
146
146
|
if (state === "done") {
|
|
147
|
-
return await downloadAndExtractMd(data.full_zip_url);
|
|
147
|
+
return cleanMarkdown(await downloadAndExtractMd(data.full_zip_url));
|
|
148
148
|
}
|
|
149
149
|
if (state === "failed") {
|
|
150
150
|
throw new Error(`MinerU Pro failed: ${data.err_msg || "unknown"}`);
|
|
@@ -202,7 +202,7 @@ async function pollBatch(
|
|
|
202
202
|
for (const r of results) {
|
|
203
203
|
if (r.state === "done" && r.full_zip_url) {
|
|
204
204
|
onProgress(`${progressPrefix} downloading ${r.file_name}…`);
|
|
205
|
-
const md = await downloadAndExtractMd(r.full_zip_url);
|
|
205
|
+
const md = cleanMarkdown(await downloadAndExtractMd(r.full_zip_url));
|
|
206
206
|
markdowns.push(md);
|
|
207
207
|
}
|
|
208
208
|
}
|
|
@@ -225,6 +225,13 @@ async function pollBatch(
|
|
|
225
225
|
throw new Error(`MinerU Pro batch ${batchId} timed out`);
|
|
226
226
|
}
|
|
227
227
|
|
|
228
|
+
// ── Output cleanup ───────────────────────────────────────────────────────────
|
|
229
|
+
|
|
230
|
+
function cleanMarkdown(md: string): string {
|
|
231
|
+
// Remove MinerU's embedded image references
|
|
232
|
+
return md.replace(/!\[.*?\]\(images\/.*?\)\n*/g, "");
|
|
233
|
+
}
|
|
234
|
+
|
|
228
235
|
// ── Public API ───────────────────────────────────────────────────────────────
|
|
229
236
|
|
|
230
237
|
export async function mineruProOcr(
|
|
@@ -245,5 +252,5 @@ export async function mineruProOcr(
|
|
|
245
252
|
const markdown = await processLocalFile(token, filePath, fileName, "[1/1]", onProgress);
|
|
246
253
|
onProgress("[1/1] done");
|
|
247
254
|
|
|
248
|
-
return { text: markdown, details: { backend: "mineru-pro", fileName } };
|
|
255
|
+
return { text: cleanMarkdown(markdown), details: { backend: "mineru-pro", fileName } };
|
|
249
256
|
}
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "pi-ocr",
|
|
3
|
-
"version": "1.3.
|
|
3
|
+
"version": "1.3.1",
|
|
4
4
|
"description": "Pi extension: Zero-setup multi-backend OCR — MinerU (free cloud), Ollama (local GPU, LaTeX formulas), Pix2Text (local Python). Extract text, formulas, and tables from images and PDFs. Default: zero config, works out of the box.",
|
|
5
5
|
"keywords": [
|
|
6
6
|
"pi-package",
|