Merge pull request #18 from jakeboone02/ignore-uoms

Add `ignoreUOMs` option

CHANGELOG.md

@@ -9,6 +9,12 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 
 - N/A
 
+## [v1.2.0] - 2024-01-18
+
+### Added
+
+- [#18] New option `ignoreUOMs`, an array of strings that will _not_ be considered units of measure.
+
 ## [v1.1.1] - 2024-01-15
 
 ### Fixed
@@ -143,10 +149,13 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 [#7]: https://github.com/jakeboone02/parse-ingredient/pull/7
 [#8]: https://github.com/jakeboone02/parse-ingredient/pull/8
 [#14]: https://github.com/jakeboone02/parse-ingredient/pull/14
+[#18]: https://github.com/jakeboone02/parse-ingredient/pull/18
 
 <!-- Release comparison links -->
 
-[unreleased]: https://github.com/jakeboone02/parse-ingredient/compare/v1.1.0...HEAD
+[unreleased]: https://github.com/jakeboone02/parse-ingredient/compare/v1.2.0...HEAD
+[v1.2.0]: https://github.com/jakeboone02/parse-ingredient/compare/v1.1.1...v1.2.0
+[v1.1.1]: https://github.com/jakeboone02/parse-ingredient/compare/v1.1.0...v1.1.1
 [v1.1.0]: https://github.com/jakeboone02/parse-ingredient/compare/v1.0.1...v1.1.0
 [v1.0.1]: https://github.com/jakeboone02/parse-ingredient/compare/v1.0.0...v1.0.1
 [v1.0.0]: https://github.com/jakeboone02/parse-ingredient/compare/v0.6.0...v1.0.0

README.md

@@ -64,7 +64,7 @@ In the browser, all exports including the `parseIngredient` function are availab
 ```html
 <script src="https://unpkg.com/parse-ingredient"></script>
 <script>
-  console.log(ParseIngredient.parseIngredient('1 1/2 cups sugar'));
+  ParseIngredient.parseIngredient('1 1/2 cups sugar');
   // [
   //   {
   //     quantity: 1.5,
@@ -83,7 +83,7 @@ In the browser, all exports including the `parseIngredient` function are availab
 ```js
 import { parseIngredient } from 'parse-ingredient';
 
-console.log(parseIngredient('1-2 pears'));
+parseIngredient('1-2 pears');
 // [
 //   {
 //     quantity: 1,
@@ -94,11 +94,10 @@ console.log(parseIngredient('1-2 pears'));
 //     isGroupHeader: false,
 //   }
 // ]
-console.log(
-  parseIngredient(
-    `2/3 cup flour
+
+parseIngredient(
+  `2/3 cup flour
 1 tsp baking powder`
-  )
 );
 // [
 //   {
@@ -118,7 +117,7 @@ console.log(
 //     isGroupHeader: false,
 //   },
 // ]
-console.log(parseIngredient('For cake:'));
+parseIngredient('For cake:');
 // [
 //   {
 //     quantity: null,
@@ -129,7 +128,7 @@ console.log(parseIngredient('For cake:'));
 //     isGroupHeader: true,
 //   }
 // ]
-console.log(parseIngredient('Ripe tomato x2'));
+parseIngredient('Ripe tomato x2');
 // [
 //   {
 //     quantity: 2,
@@ -148,10 +147,8 @@ console.log(parseIngredient('Ripe tomato x2'));
 
 Pass `true` to convert units of measure to their long, singular form, e.g. "ml" becomes "milliliter" and "cups" becomes "cup". This can help normalize the units of measure for processing. In most cases, this option will make `unitOfMeasure` equivalent to `unitOfMeasureID`.
 
-Example:
-
 ```js
-console.log(parseIngredient('1 c sugar', { normalizeUOM: true }));
+parseIngredient('1 c sugar', { normalizeUOM: true });
 // [
 //   {
 //     quantity: 1,
@@ -168,20 +165,16 @@ console.log(parseIngredient('1 c sugar', { normalizeUOM: true }));
 
 Pass an object that matches the format of the exported `unitsOfMeasure` object. Keys that match any in the exported object will be used instead of the default, and any others will be added to the list of known units of measure when parsing ingredients.
 
-Example:
-
 ```js
-console.log(
-  parseIngredient('2 buckets of widgets', {
-    additionalUOMs: {
-      bucket: {
-        short: 'bkt',
-        plural: 'buckets',
-        versions: ['bk'],
-      },
+parseIngredient('2 buckets of widgets', {
+  additionalUOMs: {
+    bucket: {
+      short: 'bkt',
+      plural: 'buckets',
+      versions: ['bk'],
     },
-  })
-);
+  },
+});
 // [
 //   {
 //     quantity: 2,
@@ -198,10 +191,8 @@ console.log(
 
 When `true`, ingredient descriptions that start with "of " will not be modified. (By default, a leading "of " will be removed from all descriptions.)
 
-Example:
-
 ```js
-console.log(parseIngredient('1 cup of sugar', { allowLeadingOf: true }));
+parseIngredient('1 cup of sugar', { allowLeadingOf: true });
 // [
 //   {
 //     quantity: 1,
@@ -214,6 +205,24 @@ console.log(parseIngredient('1 cup of sugar', { allowLeadingOf: true }));
 // ]
 ```
 
+### `ignoreUOMs`
+
+An array of strings to ignore as units of measure when parsing ingredients.
+
+```js
+parseIngredient('2 large eggs', { ignoreUOMs: ['large'] });
+// [
+//   {
+//     quantity: 2,
+//     quantity2: null,
+//     unitOfMeasure: null,
+//     unitOfMeasureID: null,
+//     description: 'large eggs',
+//     isGroupHeader: false,
+//   }
+// ]
+```
+
 ## Contributors ✨
 
 Thanks goes to these wonderful people ([emoji key](https://allcontributors.org/docs/en/emoji-key)):

ci/index.html

@@ -11,6 +11,7 @@ <h1>parse-ingredient CI</h1>
       <div>allowLeadingOf</div>
       <div>normalizeUOM</div>
       <div>additionalUOMs</div>
+      <div>ignoreUOMs</div>
     </div>
     <script src="src/index.ts"></script>
   </body>

ci/src/index.ts

@@ -20,6 +20,7 @@ const additionalUOMs = {
     alternates: ['ounce'],
   },
 } as const;
+const ignoreUOMs = ['cup', 'cups'] as const;
 
 for (const ex of examples) {
   const run = [
@@ -28,6 +29,7 @@ for (const ex of examples) {
     parseIngredient(ex, { allowLeadingOf }),
     parseIngredient(ex, { normalizeUOM }),
     parseIngredient(ex, { additionalUOMs }),
+    parseIngredient(ex, { ignoreUOMs }),
   ];
 
   gridInnerHTML.push(...run.map(e => `<div>${JSON.stringify(e, null, 2)}</div>`));

ci/src/styles.css

@@ -5,22 +5,22 @@ body {
 
 #grid {
   display: grid;
-  grid-template-columns: 1fr 1fr 1fr 1fr 1fr;
+  grid-template-columns: 1fr 1fr 1fr 1fr 1fr 1fr;
   column-gap: 1rem;
   row-gap: 0.4rem;
 }
 
-#grid > *:nth-child(n + 5) {
+#grid > *:nth-child(n + 6) {
   border-bottom: 1px solid #dddddd;
 }
 
-#grid > *:nth-child(-n + 5) {
+#grid > *:nth-child(-n + 6) {
   border-bottom: 1px solid gray;
   font-weight: bold;
   text-align: center;
 }
 
-#grid > *:nth-child(5n + 1):nth-child(n + 5) {
+#grid > *:nth-child(6n + 1):nth-child(n + 6) {
   font-weight: bold;
   text-align: right;
 }

index.html

@@ -1,4 +1,4 @@
-<!DOCTYPE html>
+<!doctype html>
 <html lang="en">
   <head>
     <meta charset="UTF-8" />
@@ -45,6 +45,11 @@ <h4>Ingredient List</h4>
       Allow leading "of "
     </label>
     <br />
+    <label>
+      Ignore UOMs
+      <input type="text" id="ignore-uoms" />
+    </label>
+    <br />
     <button type="button" id="parse">Parse</button>
     <h4>Results</h4>
     <pre id="results" class="language-json">(Click "Parse" to see results.)</pre>

src/constants.ts

@@ -8,6 +8,7 @@ export const defaultOptions = {
   additionalUOMs: {},
   allowLeadingOf: false,
   normalizeUOM: false,
+  ignoreUOMs: [],
 } as const satisfies Required<ParseIngredientOptions>;
 
 /**

src/dev.ts

@@ -26,10 +26,14 @@ document.getElementById('ingredient-list')!.innerHTML = seed;
 document.getElementById('parse')!.addEventListener('click', () => {
   const normalizeUOM = (document.getElementById('normalize-uom') as HTMLInputElement).checked;
   const allowLeadingOf = (document.getElementById('allow-leading-of') as HTMLInputElement).checked;
+  const ignoreUOMs = ((document.getElementById('ignore-uoms') as HTMLInputElement).value ?? '')
+    .split(',')
+    .map(s => s.trim());
   document.getElementById('results')!.innerHTML = JSON.stringify(
     parseIngredient((document.getElementById('ingredient-list') as HTMLInputElement).value, {
       normalizeUOM,
       allowLeadingOf,
+      ignoreUOMs,
     }),
     null,
     2

src/parseIngredient.ts

@@ -52,8 +52,11 @@ export const parseIngredient = (
       // The first character is not numeric. First check for trailing quantity/uom.
       const trailingQtyResult = trailingQuantityRegEx.exec(line);
 
-      if (trailingQtyResult) {
-        // Trailing quantity detected.
+      if (trailingQtyResult && opts.ignoreUOMs.includes(trailingQtyResult.at(-1) ?? '')) {
+        // Trailing quantity detected, but bailing out since the UOM should be ignored.
+        oIng.description = line;
+      } else if (trailingQtyResult) {
+        // Trailing quantity detected with missing or non-ignored UOM.
         // Remove the quantity and unit of measure from the description.
         oIng.description = line.replace(trailingQuantityRegEx, '').trim();
 
@@ -153,7 +156,9 @@ export const parseIngredient = (
 
         while (++i < uomArrayLength && !uom) {
           const { alternates, id, short, plural } = uomArray[i];
-          const versions = [...alternates, id, short, plural];
+          const versions = [...alternates, id, short, plural].filter(
+            unit => !opts.ignoreUOMs.includes(unit)
+          );
           if (versions.includes(firstWord)) {
             uom = firstWord;
             uomID = id;

src/parseIngredientTests.ts

@@ -526,4 +526,26 @@ export const parseIngredientTests: Record<
       },
     ],
   ],
+  'ignores units of measure': [
+    '1 cup stuff\nstuff x 2 cups',
+    [
+      {
+        quantity: 1,
+        quantity2: null,
+        unitOfMeasureID: null,
+        unitOfMeasure: null,
+        description: 'cup stuff',
+        isGroupHeader: false,
+      },
+      {
+        quantity: null,
+        quantity2: null,
+        unitOfMeasureID: null,
+        unitOfMeasure: null,
+        description: 'stuff x 2 cups',
+        isGroupHeader: false,
+      },
+    ],
+    { ignoreUOMs: ['cup', 'cups'] },
+  ],
 };

src/types.ts

@@ -69,6 +69,30 @@ export interface ParseIngredientOptions {
    * @default {}
    */
   additionalUOMs?: UnitOfMeasureDefinitions;
+  /**
+   * An array of strings to ignore as units of measure when parsing ingredients.
+   *
+   * @example
+   *
+   * ```ts
+   * parseIngredient('2 small eggs', {
+   *   ignoreUOMs: ['small', 'medium', 'large']
+   * })
+   * // [
+   * //   {
+   * //     quantity: 2,
+   * //     quantity2: null,
+   * //     unitOfMeasure: null,
+   * //     unitOfMeasureID: null,
+   * //     description: 'small eggs',
+   * //     isGroupHeader: false,
+   * //   }
+   * // ]
+   * ```
+   *
+   * @default []
+   */
+  ignoreUOMs?: string[];
   /**
    * If `true`, ingredient descriptions that start with "of " will not be
    * modified. (By default, a leading "of " will be removed from all descriptions.)