sdf-parser 6.0.1 → 7.0.3

package/README.md

@@ -1,7 +1,8 @@
 # sdf-parser
 
 [![NPM version][npm-image]][npm-url]
-[![build status][travis-image]][travis-url]
+[![build status][ci-image]][ci-url]
+[![Test coverage][codecov-image]][codecov-url]
 [![npm download][download-image]][download-url]
 
 Allow to parse a SDF file and convert it to an array of objects.
@@ -59,13 +60,19 @@ var result = parse(sdf, {
 
 ## Iterator
 
-This API is only available on Node.js.
-
 ```js
 const { iterator } = require('sdf-parser');
-const readStream = createReadStream(join(__dirname, 'test.sdf.gz'));
-const stream = readStream.pipe(createGunzip());
+const file = await openAsBlob(join(__dirname, 'test.sdf.gz'));
+
+const decompressionStream = new DecompressionStream('gzip');
+const textDecoder = new TextDecoderStream();
+
+const stream = file
+  .stream()
+  .pipeThrough(decompressionStream)
+  .pipeThrough(textDecoder);
 const results = [];
+
 for await (const entry of iterator(stream)) {
   results.push(entry);
 }
@@ -77,7 +84,9 @@ for await (const entry of iterator(stream)) {
 
 [npm-image]: https://img.shields.io/npm/v/sdf-parser.svg?style=flat-square
 [npm-url]: https://www.npmjs.com/package/sdf-parser
-[travis-image]: https://img.shields.io/travis/cheminfo/sdf-parser/main.svg?style=flat-square
-[travis-url]: https://travis-ci.org/cheminfo/sdf-parser
+[ci-image]: https://github.com/cheminfo/sdf-parser/actions/workflows/nodejs.yml/badge.svg
+[ci-url]: https://github.com/cheminfo/sdf-parser/actions/workflows/nodejs.yml
+[codecov-image]: https://img.shields.io/codecov/c/github/cheminfo/sdf-parser.svg
+[codecov-url]: https://codecov.io/gh/cheminfo/sdf-parser
 [download-image]: https://img.shields.io/npm/dm/sdf-parser.svg?style=flat-square
 [download-url]: https://www.npmjs.com/package/sdf-parser

package/lib/index.js

@@ -1,11 +1,15 @@
 'use strict';
 
-Object.defineProperty(exports, '__esModule', { value: true });
-
 var ensureString = require('ensure-string');
-var readline = require('readline');
 var dynamicTyping = require('dynamic-typing');
 
+/**
+ *
+ * @param {*} string
+ * @param {*} substring
+ * @param {*} eol
+ * @returns
+ */
 function getEntriesBoundaries(string, substring, eol) {
   const res = [];
   let previous = 0;
@@ -28,6 +32,14 @@ function getEntriesBoundaries(string, substring, eol) {
   return res;
 }
 
+/**
+ * Parse the molfile and the properties with > < labels >
+ * @param {string} sdfPart
+ * @param {*} labels
+ * @param {*} currentLabels
+ * @param {object} options
+ * @returns
+ */
 function getMolecule$1(sdfPart, labels, currentLabels, options) {
   let parts = sdfPart.split(`${options.eol}>`);
   if (parts.length === 0 || parts[0].length <= 5) return;
@@ -37,7 +49,7 @@ function getMolecule$1(sdfPart, labels, currentLabels, options) {
     let lines = parts[j].split(options.eol);
     let from = lines[0].indexOf('<');
     let to = lines[0].indexOf('>');
-    let label = lines[0].substring(from + 1, to);
+    let label = lines[0].slice(from + 1, to);
     currentLabels.push(label);
     if (!labels[label]) {
       labels[label] = {
@@ -46,8 +58,8 @@ function getMolecule$1(sdfPart, labels, currentLabels, options) {
         keep: false,
       };
       if (
-        (!options.exclude || options.exclude.indexOf(label) === -1) &&
-        (!options.include || options.include.indexOf(label) > -1)
+        (!options.exclude || !options.exclude.includes(label)) &&
+        (!options.include || options.include.includes(label))
       ) {
         labels[label].keep = true;
         if (options.modifiers[label]) {
@@ -74,10 +86,11 @@ function getMolecule$1(sdfPart, labels, currentLabels, options) {
           molecule[label] = modifiedValue;
         }
       }
-      if (labels[label].isNumeric) {
-        if (!isFinite(molecule[label]) || molecule[label].match(/^0[0-9]/)) {
-          labels[label].isNumeric = false;
-        }
+      if (
+        labels[label].isNumeric &&
+        (!Number.isFinite(+molecule[label]) || molecule[label].match(/^0[0-9]/))
+      ) {
+        labels[label].isNumeric = false;
       }
     }
   }
@@ -86,15 +99,15 @@ function getMolecule$1(sdfPart, labels, currentLabels, options) {
 
 /**
  *  Parse a SDF file
- * @param {string|ArrayBuffer|Uint8Array} sdf SDF file to parse
+ * @param {string|ArrayBuffer|Uint8Array} sdf - SDF file to parse
  * @param {object} [options={}]
- * @param {string[]} [options.include] List of fields to include
- * @param {string[]} [options.exclude] List of fields to exclude
- * @param {Function} [options.filter] Callback allowing to filter the molecules
- * @param {boolean} [options.dynamicTyping] Dynamically type the data
- * @param {object} [options.modifiers] Object containing callbacks to apply on some specific fields
- * @param {boolean} [options.mixedEOL=false] Set to true if you know there is a mixture between \r\n and \n
- * @param {string} [options.eol] Specify the end of line character. Default will be the one found in the file
+ * @param {string[]} [options.include] - List of fields to include
+ * @param {string[]} [options.exclude] - List of fields to exclude
+ * @param {Function} [options.filter] - Callback allowing to filter the molecules
+ * @param {boolean} [options.dynamicTyping] - Dynamically type the data
+ * @param {object} [options.modifiers] - Object containing callbacks to apply on some specific fields
+ * @param {boolean} [options.mixedEOL=false] - Set to true if you know there is a mixture between \r\n and \n
+ * @param {string} [options.eol] - Specify the end of line character. Default will be the one found in the file
  */
 function parse(sdf, options = {}) {
   options = { ...options };
@@ -110,14 +123,14 @@ function parse(sdf, options = {}) {
   if (options.eol === undefined) {
     options.eol = '\n';
     if (options.mixedEOL) {
-      sdf = sdf.replace(/\r\n/g, '\n');
-      sdf = sdf.replace(/\r/g, '\n');
+      sdf = sdf.replaceAll('\r\n', '\n');
+      sdf = sdf.replaceAll('\r', '\n');
     } else {
       // we will find the delimiter in order to be much faster and not use regular expression
-      let header = sdf.substr(0, 1000);
-      if (header.indexOf('\r\n') > -1) {
+      let header = new Set(sdf.slice(0, 1000));
+      if (header.has('\r\n')) {
         options.eol = '\r\n';
-      } else if (header.indexOf('\r') > -1) {
+      } else if (header.has('\r')) {
         options.eol = '\r';
       }
     }
@@ -134,7 +147,7 @@ function parse(sdf, options = {}) {
   let start = Date.now();
 
   for (let i = 0; i < entriesBoundaries.length; i++) {
-    let sdfPart = sdf.substring(...entriesBoundaries[i]);
+    let sdfPart = sdf.slice(...entriesBoundaries[i]);
 
     let currentLabels = [];
     const molecule = getMolecule$1(sdfPart, labels, currentLabels, options);
@@ -155,7 +168,7 @@ function parse(sdf, options = {}) {
       currentLabel.maxValue = -Infinity;
       for (let j = 0; j < molecules.length; j++) {
         if (molecules[j][label]) {
-          let value = parseFloat(molecules[j][label]);
+          let value = Number.parseFloat(molecules[j][label]);
           molecules[j][label] = value;
           if (value > currentLabel.maxValue) {
             currentLabel.maxValue = value;
@@ -192,57 +205,98 @@ function parse(sdf, options = {}) {
   };
 }
 
+class MolfileStream extends TransformStream {
+  #buffer = '';
+
+  constructor() {
+    super({
+      transform: (chunk, controller) => {
+        this.#buffer += chunk;
+        let begin = 0;
+        let index = 0;
+        while ((index = this.#buffer.indexOf('$$$$', index)) !== -1) {
+          // we need to check if the delimiter '\n' is in the current buffer
+          // if it is not we need to wait for the next chunk
+          const endOfDelimiter = this.#buffer.indexOf('\n', index);
+          if (endOfDelimiter === -1) {
+            index = begin;
+            break;
+          }
+          const eolLength = this.#buffer[endOfDelimiter - 1] === '\r' ? 2 : 1;
+          // need to remove the last eol because we will split on eol+'>' in getMolecule
+          controller.enqueue(this.#buffer.slice(begin, index - eolLength));
+          index = endOfDelimiter + eolLength;
+          begin = index;
+        }
+        this.#buffer = this.#buffer.slice(begin);
+      },
+      flush: (controller) => {
+        if (this.#buffer) {
+          controller.enqueue(this.#buffer);
+        }
+      },
+    });
+  }
+}
+
 /**
- *  Parse a SDF file
- * @param {NodeJS.ReadableStream} readStream SDF file to parse
- * @param {object} [options={}]
- * @param {Function} [options.filter] Callback allowing to filter the molecules
- * @param {boolean} [options.dynamicTyping] Dynamically type the data
+ *  Parse a SDF file as an iterator
+ * @param {ReadableStream} readStream - SDF file to parse
+ * @param {object} [options={}] - iterator options
+ * @param {Function} [options.filter] - Callback allowing to filter the molecules
+ * @param {string} [options.eol='\n'] - End of line character
+ * @param {boolean} [options.dynamicTyping] - Dynamically type the data
+ * @yields {object} - Molecule object
  */
-
 async function* iterator(readStream, options = {}) {
-  const lines = readline.createInterface(readStream);
-  const currentLines = [];
-  options = { ...options };
-  if (options.dynamicTyping === undefined) options.dynamicTyping = true;
+  const { eol = '\n', dynamicTyping = true } = options;
 
-  options.eol = '\n';
-  for await (let line of lines) {
-    if (line.startsWith('$$$$')) {
-      const molecule = getMolecule(currentLines.join(options.eol), options);
-      if (!options.filter || options.filter(molecule)) {
-        yield molecule;
-      }
-      currentLines.length = 0;
-    } else {
-      currentLines.push(line);
+  const moleculeStream = readStream.pipeThrough(new MolfileStream({ eol }));
+  for await (const entry of moleculeStream) {
+    if (entry.length < 20) continue;
+    const molecule = getMolecule(entry, {
+      eol,
+      dynamicTyping,
+    });
+    if (!options.filter || options.filter(molecule)) {
+      yield molecule;
     }
   }
 }
 
+/**
+ * Convert a SDF part to an object
+ * @param {string} sdfPart - text containing the molfile
+ * @param {object} options - options
+ * @param {string} options.eol - end of line character
+ * @param {boolean} options.dynamicTyping - Dynamically type the data (create numbers and booleans)
+ * @returns
+ */
 function getMolecule(sdfPart, options) {
-  let parts = sdfPart.split(`${options.eol}>`);
+  const { eol, dynamicTyping: dynamicTyping$1 } = options;
+  let parts = sdfPart.split(`${eol}>`);
   if (parts.length === 0 || parts[0].length <= 5) return;
   let molecule = {};
-  molecule.molfile = parts[0] + options.eol;
+  molecule.molfile = parts[0] + eol;
   for (let j = 1; j < parts.length; j++) {
-    let lines = parts[j].split(options.eol);
+    let lines = parts[j].split(eol);
     let from = lines[0].indexOf('<');
     let to = lines[0].indexOf('>');
-    let label = lines[0].substring(from + 1, to);
+    let label = lines[0].slice(from + 1, to);
     for (let k = 1; k < lines.length - 1; k++) {
       if (molecule[label]) {
-        molecule[label] += options.eol + lines[k];
+        molecule[label] += eol + lines[k];
       } else {
         molecule[label] = lines[k];
       }
     }
-    if (options.dynamicTyping) {
+    if (dynamicTyping$1) {
       molecule[label] = dynamicTyping.parseString(molecule[label]);
     }
   }
   return molecule;
 }
 
+exports.MolfileStream = MolfileStream;
 exports.iterator = iterator;
 exports.parse = parse;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "sdf-parser",
-  "version": "6.0.1",
+  "version": "7.0.3",
   "description": "SDF parser",
   "main": "lib/index.js",
   "module": "src/index.js",
@@ -8,9 +8,6 @@
     "lib",
     "src"
   ],
-  "browser": {
-    "./src/iterator.js": "./src/iterator.browser.js"
-  },
   "sideEffects": false,
   "scripts": {
     "build": "cheminfo-build  --entry src/index.js  --root SDFParser",
@@ -20,9 +17,8 @@
     "prepack": "npm run compile",
     "prettier": "prettier --check src",
     "prettier-write": "prettier --write src",
-    "test": "npm run test-coverage && npm run eslint",
-    "test-coverage": "jest --coverage",
-    "test-only": "jest"
+    "test": "npm run test-only && npm run eslint && npm run prettier",
+    "test-only": "vitest run --coverage"
   },
   "repository": {
     "type": "git",
@@ -43,20 +39,21 @@
   },
   "homepage": "https://github.com/cheminfo/sdf-parser",
   "devDependencies": {
-    "@babel/plugin-transform-modules-commonjs": "^7.18.6",
-    "@types/jest": "^29.1.2",
+    "@babel/plugin-transform-modules-commonjs": "^7.26.3",
+    "@types/node": "^22.13.5",
+    "@vitest/coverage-v8": "^3.0.7",
     "babel-eslint": "^10.1.0",
     "callback-stream": "^1.1.0",
-    "cheminfo-build": "^1.1.11",
-    "eslint": "^8.25.0",
-    "eslint-config-cheminfo": "^8.0.2",
-    "filelist-utils": "^1.0.1",
-    "jest": "^29.2.0",
-    "openchemlib": "^8.0.1",
-    "prettier": "^2.7.1"
+    "cheminfo-build": "^1.2.1",
+    "eslint": "^9.21.0",
+    "eslint-config-cheminfo": "^13.0.0",
+    "file-collection": "^1.0.0",
+    "openchemlib": "^8.18.1",
+    "prettier": "^3.5.2",
+    "vitest": "^3.0.7"
   },
   "dependencies": {
-    "dynamic-typing": "^1.0.0",
+    "dynamic-typing": "^1.0.1",
     "ensure-string": "^1.2.0"
   }
 }

package/src/MolfileStream.js

@@ -0,0 +1,33 @@
+export class MolfileStream extends TransformStream {
+  #buffer = '';
+
+  constructor() {
+    super({
+      transform: (chunk, controller) => {
+        this.#buffer += chunk;
+        let begin = 0;
+        let index = 0;
+        while ((index = this.#buffer.indexOf('$$$$', index)) !== -1) {
+          // we need to check if the delimiter '\n' is in the current buffer
+          // if it is not we need to wait for the next chunk
+          const endOfDelimiter = this.#buffer.indexOf('\n', index);
+          if (endOfDelimiter === -1) {
+            index = begin;
+            break;
+          }
+          const eolLength = this.#buffer[endOfDelimiter - 1] === '\r' ? 2 : 1;
+          // need to remove the last eol because we will split on eol+'>' in getMolecule
+          controller.enqueue(this.#buffer.slice(begin, index - eolLength));
+          index = endOfDelimiter + eolLength;
+          begin = index;
+        }
+        this.#buffer = this.#buffer.slice(begin);
+      },
+      flush: (controller) => {
+        if (this.#buffer) {
+          controller.enqueue(this.#buffer);
+        }
+      },
+    });
+  }
+}

package/src/getEntriesBoundaries.js

@@ -1,3 +1,10 @@
+/**
+ *
+ * @param {*} string
+ * @param {*} substring
+ * @param {*} eol
+ * @returns
+ */
 export function getEntriesBoundaries(string, substring, eol) {
   const res = [];
   let previous = 0;

package/src/index.js

@@ -1,2 +1,3 @@
 export * from './parse';
 export * from './iterator';
+export * from './MolfileStream';

package/src/iterator.js

@@ -1,52 +1,59 @@
-import { createInterface } from 'readline';
-
 import { parseString } from 'dynamic-typing';
+
+import { MolfileStream } from './MolfileStream.js';
+
 /**
- *  Parse a SDF file
- * @param {NodeJS.ReadableStream} readStream SDF file to parse
- * @param {object} [options={}]
- * @param {Function} [options.filter] Callback allowing to filter the molecules
- * @param {boolean} [options.dynamicTyping] Dynamically type the data
+ *  Parse a SDF file as an iterator
+ * @param {ReadableStream} readStream - SDF file to parse
+ * @param {object} [options={}] - iterator options
+ * @param {Function} [options.filter] - Callback allowing to filter the molecules
+ * @param {string} [options.eol='\n'] - End of line character
+ * @param {boolean} [options.dynamicTyping] - Dynamically type the data
+ * @yields {object} - Molecule object
  */
-
 export async function* iterator(readStream, options = {}) {
-  const lines = createInterface(readStream);
-  const currentLines = [];
-  options = { ...options };
-  if (options.dynamicTyping === undefined) options.dynamicTyping = true;
+  const { eol = '\n', dynamicTyping = true } = options;
 
-  options.eol = '\n';
-  for await (let line of lines) {
-    if (line.startsWith('$$$$')) {
-      const molecule = getMolecule(currentLines.join(options.eol), options);
-      if (!options.filter || options.filter(molecule)) {
-        yield molecule;
-      }
-      currentLines.length = 0;
-    } else {
-      currentLines.push(line);
+  const moleculeStream = readStream.pipeThrough(new MolfileStream({ eol }));
+  for await (const entry of moleculeStream) {
+    if (entry.length < 20) continue;
+    const molecule = getMolecule(entry, {
+      eol,
+      dynamicTyping,
+    });
+    if (!options.filter || options.filter(molecule)) {
+      yield molecule;
     }
   }
 }
 
+/**
+ * Convert a SDF part to an object
+ * @param {string} sdfPart - text containing the molfile
+ * @param {object} options - options
+ * @param {string} options.eol - end of line character
+ * @param {boolean} options.dynamicTyping - Dynamically type the data (create numbers and booleans)
+ * @returns
+ */
 function getMolecule(sdfPart, options) {
-  let parts = sdfPart.split(`${options.eol}>`);
+  const { eol, dynamicTyping } = options;
+  let parts = sdfPart.split(`${eol}>`);
   if (parts.length === 0 || parts[0].length <= 5) return;
   let molecule = {};
-  molecule.molfile = parts[0] + options.eol;
+  molecule.molfile = parts[0] + eol;
   for (let j = 1; j < parts.length; j++) {
-    let lines = parts[j].split(options.eol);
+    let lines = parts[j].split(eol);
     let from = lines[0].indexOf('<');
     let to = lines[0].indexOf('>');
-    let label = lines[0].substring(from + 1, to);
+    let label = lines[0].slice(from + 1, to);
     for (let k = 1; k < lines.length - 1; k++) {
       if (molecule[label]) {
-        molecule[label] += options.eol + lines[k];
+        molecule[label] += eol + lines[k];
       } else {
         molecule[label] = lines[k];
       }
     }
-    if (options.dynamicTyping) {
+    if (dynamicTyping) {
       molecule[label] = parseString(molecule[label]);
     }
   }

package/src/parse.js

@@ -4,15 +4,15 @@ import { getEntriesBoundaries } from './getEntriesBoundaries';
 import { getMolecule } from './util/getMolecule';
 /**
  *  Parse a SDF file
- * @param {string|ArrayBuffer|Uint8Array} sdf SDF file to parse
+ * @param {string|ArrayBuffer|Uint8Array} sdf - SDF file to parse
  * @param {object} [options={}]
- * @param {string[]} [options.include] List of fields to include
- * @param {string[]} [options.exclude] List of fields to exclude
- * @param {Function} [options.filter] Callback allowing to filter the molecules
- * @param {boolean} [options.dynamicTyping] Dynamically type the data
- * @param {object} [options.modifiers] Object containing callbacks to apply on some specific fields
- * @param {boolean} [options.mixedEOL=false] Set to true if you know there is a mixture between \r\n and \n
- * @param {string} [options.eol] Specify the end of line character. Default will be the one found in the file
+ * @param {string[]} [options.include] - List of fields to include
+ * @param {string[]} [options.exclude] - List of fields to exclude
+ * @param {Function} [options.filter] - Callback allowing to filter the molecules
+ * @param {boolean} [options.dynamicTyping] - Dynamically type the data
+ * @param {object} [options.modifiers] - Object containing callbacks to apply on some specific fields
+ * @param {boolean} [options.mixedEOL=false] - Set to true if you know there is a mixture between \r\n and \n
+ * @param {string} [options.eol] - Specify the end of line character. Default will be the one found in the file
  */
 export function parse(sdf, options = {}) {
   options = { ...options };
@@ -28,14 +28,14 @@ export function parse(sdf, options = {}) {
   if (options.eol === undefined) {
     options.eol = '\n';
     if (options.mixedEOL) {
-      sdf = sdf.replace(/\r\n/g, '\n');
-      sdf = sdf.replace(/\r/g, '\n');
+      sdf = sdf.replaceAll('\r\n', '\n');
+      sdf = sdf.replaceAll('\r', '\n');
     } else {
       // we will find the delimiter in order to be much faster and not use regular expression
-      let header = sdf.substr(0, 1000);
-      if (header.indexOf('\r\n') > -1) {
+      let header = new Set(sdf.slice(0, 1000));
+      if (header.has('\r\n')) {
         options.eol = '\r\n';
-      } else if (header.indexOf('\r') > -1) {
+      } else if (header.has('\r')) {
         options.eol = '\r';
       }
     }
@@ -52,7 +52,7 @@ export function parse(sdf, options = {}) {
   let start = Date.now();
 
   for (let i = 0; i < entriesBoundaries.length; i++) {
-    let sdfPart = sdf.substring(...entriesBoundaries[i]);
+    let sdfPart = sdf.slice(...entriesBoundaries[i]);
 
     let currentLabels = [];
     const molecule = getMolecule(sdfPart, labels, currentLabels, options);
@@ -73,7 +73,7 @@ export function parse(sdf, options = {}) {
       currentLabel.maxValue = -Infinity;
       for (let j = 0; j < molecules.length; j++) {
         if (molecules[j][label]) {
-          let value = parseFloat(molecules[j][label]);
+          let value = Number.parseFloat(molecules[j][label]);
           molecules[j][label] = value;
           if (value > currentLabel.maxValue) {
             currentLabel.maxValue = value;

package/src/util/getMolecule.js

@@ -1,3 +1,11 @@
+/**
+ * Parse the molfile and the properties with > < labels >
+ * @param {string} sdfPart
+ * @param {*} labels
+ * @param {*} currentLabels
+ * @param {object} options
+ * @returns
+ */
 export function getMolecule(sdfPart, labels, currentLabels, options) {
   let parts = sdfPart.split(`${options.eol}>`);
   if (parts.length === 0 || parts[0].length <= 5) return;
@@ -7,7 +15,7 @@ export function getMolecule(sdfPart, labels, currentLabels, options) {
     let lines = parts[j].split(options.eol);
     let from = lines[0].indexOf('<');
     let to = lines[0].indexOf('>');
-    let label = lines[0].substring(from + 1, to);
+    let label = lines[0].slice(from + 1, to);
     currentLabels.push(label);
     if (!labels[label]) {
       labels[label] = {
@@ -16,8 +24,8 @@ export function getMolecule(sdfPart, labels, currentLabels, options) {
         keep: false,
       };
       if (
-        (!options.exclude || options.exclude.indexOf(label) === -1) &&
-        (!options.include || options.include.indexOf(label) > -1)
+        (!options.exclude || !options.exclude.includes(label)) &&
+        (!options.include || options.include.includes(label))
       ) {
         labels[label].keep = true;
         if (options.modifiers[label]) {
@@ -44,10 +52,11 @@ export function getMolecule(sdfPart, labels, currentLabels, options) {
           molecule[label] = modifiedValue;
         }
       }
-      if (labels[label].isNumeric) {
-        if (!isFinite(molecule[label]) || molecule[label].match(/^0[0-9]/)) {
-          labels[label].isNumeric = false;
-        }
+      if (
+        labels[label].isNumeric &&
+        (!Number.isFinite(+molecule[label]) || molecule[label].match(/^0[0-9]/))
+      ) {
+        labels[label].isNumeric = false;
       }
     }
   }

package/src/__tests__/__snapshots__/getEntriesBoundaries.test.js.snap

@@ -1,10 +0,0 @@
-// Jest Snapshot v1, https://goo.gl/fbAQLP
-
-exports[`should parse sdf files without EOL in the EOF 1`] = `
-[
-  [
-    0,
-    1473,
-  ],
-]
-`;