feat: add ocr function

package.json

@@ -27,6 +27,7 @@
     "quasar": "^2.16.9",
     "signature_pad": "^5.0.2",
     "socket.io-client": "^4.7.5",
+    "tesseract.js": "^5.1.1",
     "uuid": "^10.0.0",
     "vue": "^3.4.38",
     "vue-i18n": "^9.14.0",

pnpm-lock.yaml

@@ -47,6 +47,9 @@ importers:
       socket.io-client:
         specifier: ^4.7.5
         version: 4.7.5
+      tesseract.js:
+        specifier: ^5.1.1
+        version: 5.1.1
       uuid:
         specifier: ^10.0.0
         version: 10.0.0
@@ -1140,6 +1143,9 @@ packages:
   bluebird@3.7.2:
     resolution: {integrity: sha512-XpNj6GDQzdfW+r2Wnn7xiSAd7TM3jzkxGXBGTtWKuSXv1xUV+azxAm8jdWZN06QTQk+2N2XB9jRDkvbmQmcRtg==}
 
+  bmp-js@0.1.0:
+    resolution: {integrity: sha512-vHdS19CnY3hwiNdkaqk93DvjVLfbEcI8mys4UjuWrlX1haDmroo8o4xCzh4wD6DGV6HxRCyauwhHRqMTfERtjw==}
+
   bn.js@4.12.0:
     resolution: {integrity: sha512-c98Bf3tPniI+scsdk237ku1Dc3ujXQTSgyiPUDEOe7tRkhrqridvh8klBv0HCEso1OLOYcHuCv/cS6DNxKH+ZA==}
 
@@ -1262,7 +1268,6 @@ packages:
 
   chokidar@2.1.8:
     resolution: {integrity: sha512-ZmZUazfOzf0Nve7duiCKD23PFSCs4JPoYyccjUFF3aQkQadqBhfzhjkwBH2mNOG9cTBwhamM37EIsIkZw3nRgg==}
-    deprecated: Chokidar 2 does not receive security updates since 2019. Upgrade to chokidar 3 with 15x fewer dependencies
 
   chokidar@3.6.0:
     resolution: {integrity: sha512-7VT13fmjotKpGipCW9JEQAusEPE+Ei8nl6/g4FBAmIm0GOOLMua9NDDo/DWp0ZAxCr3cPq5ZpBqmPAQgDda2Pw==}
@@ -1880,7 +1885,7 @@ packages:
     resolution: {integrity: sha512-oWb1Z6mkHIskLzEJ/XWX0srkpkTQ7vaopMQkyaEIoq0fmtFVxOthb8cCxeT+p3ynTdkk/RZwbgG4brR5BeWECw==}
     engines: {node: '>= 4.0'}
     os: [darwin]
-    deprecated: The v1 package contains DANGEROUS / INSECURE binaries. Upgrade to safe fsevents v2
+    deprecated: Upgrade to fsevents v2 to mitigate potential security issues
 
   fsevents@2.3.2:
     resolution: {integrity: sha512-xiqMQR4xAeHTuB9uWm+fFRcIOgKBMiOBP+eXiyT7jsgVCq1bkVygt00oASowB7EdtpOHaaPgKt812P9ab+DDKA==}
@@ -2019,6 +2024,9 @@ packages:
     resolution: {integrity: sha512-v3MXnZAcvnywkTUEZomIActle7RXXeedOR31wwl7VlyoXO4Qi9arvSenNQWne1TcRwhCL1HwLI21bEqdpj8/rA==}
     engines: {node: '>=0.10.0'}
 
+  idb-keyval@6.2.1:
+    resolution: {integrity: sha512-8Sb3veuYCyrZL+VBt9LJfZjLUPWVvqn8tG28VqYNFCo43KHcKuq+b4EiXGeuaLAQWL2YmyDgMp2aSpH9JHsEQg==}
+
   ieee754@1.2.1:
     resolution: {integrity: sha512-dcyqhDvX1C46lXZcVqCpK+FtMRQVdIMN6/Df5js2zouUsqG7I6sFxitIC+7KYK29KdXOLHdu9zL4sFnoVQnqaA==}
 
@@ -2097,6 +2105,9 @@ packages:
     engines: {node: ^12.20.0 || ^14.13.1 || >=16.0.0}
     hasBin: true
 
+  is-electron@2.2.2:
+    resolution: {integrity: sha512-FO/Rhvz5tuw4MCWkpMzHFKWD2LsfHzIb7i6MdPYZ/KW7AlxawyLkqdy+jPZP1WubqEADE3O4FUENlJHDfQASRg==}
+
   is-extendable@0.1.1:
     resolution: {integrity: sha512-5BMULNob1vgFX6EjQw5izWDxrecWK9AM72rugNr0TFldMOi0fj6Jk+zeKIt0xGj4cEfQIJth4w3OKWOJ4f+AFw==}
     engines: {node: '>=0.10.0'}
@@ -2154,6 +2165,9 @@ packages:
     resolution: {integrity: sha512-knxG2q4UC3u8stRGyAVJCOdxFmv5DZiRcdlIaAQXAbSfJya+OhopNotLQrstBhququ4ZpuKbDc/8S6mgXgPFPw==}
     engines: {node: '>=10'}
 
+  is-url@1.2.4:
+    resolution: {integrity: sha512-ITvGim8FhRiYe4IQ5uHSkj7pVaPDrCTkNd3yq3cV7iZAcJdHTUMPMEHcqSOy9xZ9qFenQCvi+2wjH9a1nXqHww==}
+
   is-windows@1.0.2:
     resolution: {integrity: sha512-eXK1UInq2bPmjyX6e3VHIzMLobc4J94i4AWn+Hpq3OU5KkrRC96OAcR3PRJ/pGu6m8TRnBHP9dkXQVsT/COVIA==}
     engines: {node: '>=0.10.0'}
@@ -2594,6 +2608,10 @@ packages:
     resolution: {integrity: sha512-7x81NCL719oNbsq/3mh+hVrAWmFuEYUqrq/Iw3kUzH8ReypT9QQ0BLoJS7/G9k6N81XjW4qHWtjWwe/9eLy1EQ==}
     engines: {node: '>=12'}
 
+  opencollective-postinstall@2.0.3:
+    resolution: {integrity: sha512-8AV/sCtuzUeTo8gQK5qDZzARrulB3egtLzFgteqB2tcT4Mw7B8Kt7JcDHmltjz6FOAHsvTevk70gZEbhM4ZS9Q==}
+    hasBin: true
+
   optionator@0.9.3:
     resolution: {integrity: sha512-JjCoypp+jKn1ttEFExxhetCKeJt9zhAgAve5FXHixTvFDW/5aEktX9bufBKLRRMdU7bNtpLfcGu94B3cdEJgjg==}
     engines: {node: '>= 0.8.0'}
@@ -2890,6 +2908,9 @@ packages:
     resolution: {integrity: sha512-hOS089on8RduqdbhvQ5Z37A0ESjsqz6qnRcffsMU3495FuTdqSm+7bhJ29JvIOsBDEEnan5DPu9t3To9VRlMzA==}
     engines: {node: '>=8.10.0'}
 
+  regenerator-runtime@0.13.11:
+    resolution: {integrity: sha512-kY1AZVr2Ra+t+piVaJ4gxaFaReZVH40AKNo7UCX6W+dEwBo/2oZJzqfuN1qLq1oL45o56cPaTXELwrTh8Fpggg==}
+
   regex-not@1.0.2:
     resolution: {integrity: sha512-J6SDjUgDxQj5NusnOtdFxDwN/+HWykR8GELwctJ7mdqhcyy1xEc4SRFHUXvxTp661YaVKAjfRLZ9cCqS6tn32A==}
     engines: {node: '>=0.10.0'}
@@ -3255,6 +3276,12 @@ packages:
     engines: {node: '>=10'}
     hasBin: true
 
+  tesseract.js-core@5.1.1:
+    resolution: {integrity: sha512-KX3bYSU5iGcO1XJa+QGPbi+Zjo2qq6eBhNjSGR5E5q0JtzkoipJKOUQD7ph8kFyteCEfEQ0maWLu8MCXtvX5uQ==}
+
+  tesseract.js@5.1.1:
+    resolution: {integrity: sha512-lzVl/Ar3P3zhpUT31NjqeCo1f+D5+YfpZ5J62eo2S14QNVOmHBTtbchHm/YAbOOOzCegFnKf4B3Qih9LuldcYQ==}
+
   text-decoder@1.1.1:
     resolution: {integrity: sha512-8zll7REEv4GDD3x4/0pW+ppIxSNs7H1J10IKFZsuOMscumCdM2a+toDGLPA3T+1+fLBql4zbt5z83GEQGGV5VA==}
 
@@ -3504,6 +3531,9 @@ packages:
       typescript:
         optional: true
 
+  wasm-feature-detect@1.8.0:
+    resolution: {integrity: sha512-zksaLKM2fVlnB5jQQDqKXXwYHLQUVH9es+5TOOHwGOVJOCeRBCiPjwSg+3tN2AdTCzjgli4jijCH290kXb/zWQ==}
+
   watchpack-chokidar2@2.0.1:
     resolution: {integrity: sha512-nCFfBIPKr5Sh61s4LPpy1Wtfi0HE8isJ3d2Yb5/Ppw2P2B/3eVSEBjKfN0fmHJSK14+31KwMKmcrzs2GM4P0Ww==}
 
@@ -3647,6 +3677,9 @@ packages:
     resolution: {integrity: sha512-zK7YHHz4ZXpW89AHXUPbQVGKI7uvkd3hzusTdotCg1UxyaVtg0zFJSTfW/Dq5f7OBBVnq6cZIaC8Ti4hb6dtCA==}
     engines: {node: '>= 14'}
 
+  zlibjs@0.3.1:
+    resolution: {integrity: sha512-+J9RrgTKOmlxFSDHo0pI1xM6BLVUv+o0ZT9ANtCxGkjIVCCUdx9alUF8Gm+dGLKbkkkidWIHFDZHDMpfITt4+w==}
+
 snapshots:
 
   '@aashutoshrathi/word-wrap@1.2.6': {}
@@ -4670,6 +4703,8 @@ snapshots:
 
   bluebird@3.7.2: {}
 
+  bmp-js@0.1.0: {}
+
   bn.js@4.12.0: {}
 
   bn.js@5.2.1: {}
@@ -5837,6 +5872,8 @@ snapshots:
     dependencies:
       safer-buffer: 2.1.2
 
+  idb-keyval@6.2.1: {}
+
   ieee754@1.2.1: {}
 
   iferr@0.1.5: {}
@@ -5913,6 +5950,8 @@ snapshots:
 
   is-docker@3.0.0: {}
 
+  is-electron@2.2.2: {}
+
   is-extendable@0.1.1: {}
 
   is-extendable@1.0.1:
@@ -5954,6 +5993,8 @@ snapshots:
 
   is-unicode-supported@0.1.0: {}
 
+  is-url@1.2.4: {}
+
   is-windows@1.0.2: {}
 
   is-wsl@1.1.0: {}
@@ -6312,7 +6353,6 @@ snapshots:
   node-fetch@2.7.0:
     dependencies:
       whatwg-url: 5.0.0
-    optional: true
 
   node-forge@1.3.1: {}
 
@@ -6427,6 +6467,8 @@ snapshots:
       is-docker: 2.2.1
       is-wsl: 2.2.0
 
+  opencollective-postinstall@2.0.3: {}
+
   optionator@0.9.3:
     dependencies:
       '@aashutoshrathi/word-wrap': 1.2.6
@@ -6738,6 +6780,8 @@ snapshots:
     dependencies:
       picomatch: 2.3.1
 
+  regenerator-runtime@0.13.11: {}
+
   regex-not@1.0.2:
     dependencies:
       extend-shallow: 3.0.2
@@ -7182,6 +7226,23 @@ snapshots:
       commander: 2.20.3
       source-map-support: 0.5.21
 
+  tesseract.js-core@5.1.1: {}
+
+  tesseract.js@5.1.1:
+    dependencies:
+      bmp-js: 0.1.0
+      idb-keyval: 6.2.1
+      is-electron: 2.2.2
+      is-url: 1.2.4
+      node-fetch: 2.7.0
+      opencollective-postinstall: 2.0.3
+      regenerator-runtime: 0.13.11
+      tesseract.js-core: 5.1.1
+      wasm-feature-detect: 1.8.0
+      zlibjs: 0.3.1
+    transitivePeerDependencies:
+      - encoding
+
   text-decoder@1.1.1:
     dependencies:
       b4a: 1.6.6
@@ -7227,8 +7288,7 @@ snapshots:
 
   toidentifier@1.0.1: {}
 
-  tr46@0.0.3:
+  tr46@0.0.3: {}
-    optional: true
 
   ts-api-utils@1.3.0(typescript@5.5.4):
     dependencies:
@@ -7399,6 +7459,8 @@ snapshots:
     optionalDependencies:
       typescript: 5.5.4
 
+  wasm-feature-detect@1.8.0: {}
+
   watchpack-chokidar2@2.0.1:
     dependencies:
       chokidar: 2.1.8
@@ -7420,8 +7482,7 @@ snapshots:
     dependencies:
       defaults: 1.0.4
 
-  webidl-conversions@3.0.1:
+  webidl-conversions@3.0.1: {}
-    optional: true
 
   webpack-merge@6.0.1:
     dependencies:
@@ -7470,7 +7531,6 @@ snapshots:
     dependencies:
       tr46: 0.0.3
       webidl-conversions: 3.0.1
-    optional: true
 
   which@2.0.2:
     dependencies:
@@ -7558,3 +7618,5 @@ snapshots:
       archiver-utils: 5.0.2
       compress-commons: 6.0.2
       readable-stream: 4.5.2
+
+  zlibjs@0.3.1: {}

src/utils/mrz.ts

@@ -0,0 +1,135 @@
+type MRZ = {
+  type: 'TD1' | 'TD2' | 'TD3';
+  zone: string[];
+};
+
+const MRZ_TD_1 = [
+  new RegExp(
+    [
+      '(?<doc_type>[0-9A-Z<]{1})',
+      '(?<doc_subtype>[A-Z<]{1})',
+      '(?<country>[0-9A-Z<]{3})',
+      '(?<doc_number>[0-9A-Z<]{9})',
+      '(?<doc_number_check>[0-9A-Z<]{1})',
+      '(?<complement>[0-9A-Z<]{15})',
+    ].join(''),
+  ),
+  new RegExp(
+    [
+      '(?<birth_date>[0-9A-Z<]{6})',
+      '(?<birth_date_check>[0-9A-Z<]{1})',
+      '(?<sex>[mfMF<]{1})',
+      '(?<expire_date>[0-9A-Z<]{6})',
+      '(?<expire_date_check>[0-9A-Z<]{1})',
+      '(?<nationality>[0-9A-Z<]{3})',
+      '(?<optional_data>[A-Z0-9<]{11})',
+      '(?<linecheck>[0-9A-Z<]{1})',
+    ].join(''),
+  ),
+  new RegExp(['(?<full_name>[A-Z<]{30})'].join('')),
+];
+const MRZ_TD_2 = [
+  new RegExp(
+    [
+      '(?<doc_type>[0-9A-Z<]{1})',
+      '(?<doc_subtype>[A-Z<]{1})',
+      '(?<country>[0-9A-Z<]{3})',
+      '(?<full_name>[A-Z<]{31})',
+    ].join(''),
+  ),
+
+  new RegExp(
+    [
+      '(?<doc_number>[0-9A-Z<]{9})',
+      '(?<doc_numbercheck>[0-9A-Z<]{1})',
+      '(?<nacionality>[0-9A-Z<]{3})',
+      '(?<birth_date>[0-9A-Z<]{6})',
+      '(?<birth_date_check>[0-9A-Z<]{1})',
+      '(?<sex>[mfMF]{1})',
+      '(?<expire_date>[0-9A-Z<]{6})',
+      '(?<expire_date_check>[0-9A-Z<]{1})',
+      '(?<optional_data>[A-Z0-9<]{7})',
+      '(?<line_check>[0-9A-Z<]{1})',
+    ].join(''),
+  ),
+];
+
+const MRZ_TD_3 = [
+  new RegExp(
+    [
+      '(?<doc_type>[A-Z0-9<]{1})',
+      '(?<doc_subtype>[A-Z0-9<]{1})',
+      '(?<country>[A-Z0-9]{3})',
+      '(?<full_name>[A-Z0-9<]{39})',
+    ].join(''),
+  ),
+
+  new RegExp(
+    [
+      '(?<doc_number>[0-9A-Z<]{9})',
+      '(?<doc_number_check>[0-9A-Z<]{1})',
+      '(?<nationality>[0-9A-Z<]{3})',
+      '(?<birth_date>[0-9A-Z<]{6})',
+      '(?<birth_date_check>[0-9A-Z<]{1})',
+      '(?<sex>[mfMF<]{1})',
+      '(?<expire_date>[0-9A-Z<]{6})',
+      '(?<expire_date_check>[0-9A-Z<]{1})',
+      '(?<personal_number>[A-Z0-9<]{14})',
+      '(?<personal_number_check>[0-9A-Z<]{1})',
+      '(?<linecheck>[0-9A-Z<]{1})',
+    ].join(''),
+  ),
+];
+
+function mrzCleanResult(obj: Record<string, string>) {
+  Object.entries(obj).forEach(([k, v]) => {
+    obj[k] = v
+      .replace(/</g, ' ')
+      .replace(/\s{2,}/, ' ')
+      .trim();
+  });
+  return obj;
+}
+
+export function parseType1(mrz: MRZ) {
+  const result: Record<string, string> = {};
+
+  mrz.zone.forEach((line, i) => {
+    if (MRZ_TD_1[i].test(line)) {
+      Object.assign(result, MRZ_TD_1[i].exec(line)?.groups);
+    }
+  });
+
+  return { mrz, result: mrzCleanResult(result) };
+}
+
+export function parseType2(mrz: MRZ) {
+  const result: Record<string, string> = {};
+
+  mrz.zone.forEach((line, i) => {
+    if (MRZ_TD_2[i].test(line)) {
+      Object.assign(result, MRZ_TD_2[i].exec(line)?.groups);
+    }
+  });
+
+  return { mrz, result: mrzCleanResult(result) };
+}
+
+export function parseType3(mrz: MRZ) {
+  const result: Record<string, string> = {};
+
+  mrz.zone.forEach((line, i) => {
+    if (MRZ_TD_3[i].test(line)) {
+      Object.assign(result, MRZ_TD_3[i].exec(line)?.groups);
+    }
+  });
+
+  return { mrz, result: mrzCleanResult(result) };
+}
+
+export function parseMRZ(mrz: MRZ) {
+  if (mrz.type === 'TD1') return parseType1(mrz);
+  if (mrz.type === 'TD2') return parseType2(mrz);
+  if (mrz.type === 'TD3') return parseType3(mrz);
+  return null;
+}

src/utils/ocr.ts

@@ -0,0 +1,38 @@
+import { createWorker, ImageLike, RecognizeResult } from 'tesseract.js';
+
+import { parseMRZ } from './mrz';
+
+let worker: Tesseract.Worker;
+
+export function runOcr(image: ImageLike): Promise<void>;
+export function runOcr<T extends (result: RecognizeResult) => void>(
+  image: ImageLike,
+  callback: T,
+): Promise<ReturnType<T>>;
+export async function runOcr<T extends (result: RecognizeResult) => void>(
+  image: ImageLike,
+  callback?: T,
+) {
+  if (!worker) {
+    worker = await createWorker(['ocrb', 'eng', 'khm'], 1, {
+      langPath: '/ocr-data',
+    });
+  }
+  if (callback) return callback(await worker.recognize(image));
+}
+
+export function parseResultMRZ(result: RecognizeResult) {
+  const zone = result.data.text
+    .split(/[\s\r\n]+/)
+    .filter((v) => /[A-Z0-9<]{30,}/.test(v));
+
+  if (zone.length === 3 && zone[0].length === 30) {
+    return parseMRZ({ type: 'TD1', zone });
+  } else if (zone.length === 2 && zone[0].length === 36) {
+    return parseMRZ({ type: 'TD2', zone });
+  } else if (zone.length === 2 && zone[0].length === 44) {
+    return parseMRZ({ type: 'TD3', zone });
+  } else {
+    return null;
+  }
+}