polytypo 1.12.0 → 1.14.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8c8aba5b27eb6b2b852217b3c4ba1652d78fee2fd291ce2a14e8cfc36da9ea5f
-  data.tar.gz: 321cf32a34decf2b526b6ae4231df91cf8a7d07c96b55a211b25582e701c9abe
+  metadata.gz: 40f4b8a41c44a06a195e83cbc9d8928d82b6ae5ae314c22e263063451fee27d8
+  data.tar.gz: f1a9c674f4bb0fac75872802b3ede5c10063a85cb4c80919a737aa6f49a2c640
 SHA512:
-  metadata.gz: d4ba1cd320aeaca4076180463c47d0d802fa3c0c9672a34326bbef2cd39ce6308b1209454792eccc2a534aec8476c840413b9f4cdddbd8cf248ef0f7205c1f0f
-  data.tar.gz: f5dd45b49cec16dcfdca6994646d6083242cfc8cb1357ec687786bfb73f24390af6b940fe49570ab0183d6507c79c4bb29581c5fdd40caa208a4bc2d3666845d
+  metadata.gz: d29173f633da9a14abffef5db46f2052fdd5928429b35192865cf677aa3503d0e8b836f2eb9d66dea9a8a4d8b315008ff2367673d3992e98c6e87c012f26a8e5
+  data.tar.gz: 37aeef545d793fb1a99659c7fc1d2f9ebfce3c24800a16060d3d3f4597dccd8b091fe3d2097ca2008bd9a0675730fe380f5d5732ed61f6a904ab1e91645c7baa

data/README.md

@@ -110,6 +110,26 @@ the `''` is left alone and the text on either side is processed. The block is al
 the body, which shows up in exactly one place: an unbalanced quotation mark in `title` cannot pair
 with a mark in your first paragraph.
 
+### Protecting regions of a text
+
+In `text` mode, `protect:` names ranges of the input that are never edited and that the rules read
+as opaque content, which is what a template hole is: `${…}` in a JavaScript template literal, an
+ICU argument, a `{{ … }}` tag. Ranges are `[start, end]` pairs of code-point offsets (not bytes and
+not UTF-16 units), half-open, ascending and non-overlapping; touching ranges are one run.
+
+```ruby
+Polytypo.transform("${x} - Rules", locale: "en-US", protect: [[0, 4]])
+# => "${x}—Rules"
+Polytypo.transform("${x} - Rules", locale: "en-GB", protect: [[0, 4]])
+# => "${x} – Rules"
+```
+
+The protected code points come back exactly as written. Edits before a range shift it, so a second
+call over the output needs the ranges at their new offsets; `transform` does not return them.
+`nil` and `[]` mean no protection. A malformed range, or a non-empty list in `html`, `markdown` or
+`yaml` mode, raises `Polytypo::Error` with `POLYTYPO_INVALID_OPTION`. `Polytypo.analyze` takes the
+same option and reports offsets into the input as passed.
+
 `require "polytypo"` never loads the native `commonmarker` extension unless you actually call
 `Polytypo.transform` with `mode: "markdown"` — the require happens lazily, inside the markdown
 pipeline only. A single `Polytypo.transform` module method (not separate
@@ -135,7 +155,7 @@ no-break one back), so replaying the list is not guaranteed to reproduce the out
 
 ### Errors
 
-Every error `Polytypo.transform` raises is a `Polytypo::Error` carrying one of seven stable
+Every error `Polytypo.transform` raises is a `Polytypo::Error` carrying one of eight stable
 codes — check `#code`, not the message text, which is English and informative but not part of
 the contract:
 

data/lib/polytypo/data/VERSION

@@ -1 +1 @@
-1.12.0
+1.14.0

data/lib/polytypo/data/fixtures/bg.json

@@ -1,5 +1,5 @@
 {
-  "spec": "1.12.0",
+  "spec": "1.14.0",
   "locale": "bg",
   "cases": [
     {

data/lib/polytypo/data/fixtures/cs.json

@@ -1,5 +1,5 @@
 {
-  "spec": "1.12.0",
+  "spec": "1.14.0",
   "locale": "cs",
   "cases": [
     {

data/lib/polytypo/data/fixtures/de-CH.json

@@ -1,5 +1,5 @@
 {
-  "spec": "1.12.0",
+  "spec": "1.14.0",
   "locale": "de-CH",
   "cases": [
     {

data/lib/polytypo/data/fixtures/de-DE.json

@@ -1,5 +1,5 @@
 {
-  "spec": "1.12.0",
+  "spec": "1.14.0",
   "locale": "de-DE",
   "cases": [
     {
@@ -787,6 +787,82 @@
       "in": "description: 'a''--''b'\n",
       "out": "description: 'a''--''b'\n",
       "note": "modes.md §3.8.6 and §3.4: de-DE is a `-spaced` locale, so the dash would grow against a unit and the edge-growth rule discards it, as in fr. A second locale so that an implementation that does not split cannot pass on one cell alone."
+    },
+    {
+      "id": "de-de-protect-two-holes-dash",
+      "rule": "dashes",
+      "mode": "text",
+      "in": "${a} - ${b}",
+      "protect": [
+        {
+          "start": 0,
+          "end": 4
+        },
+        {
+          "start": 7,
+          "end": 11
+        }
+      ],
+      "out": "${a} – ${b}",
+      "protectOut": [
+        {
+          "start": 0,
+          "end": 4
+        },
+        {
+          "start": 7,
+          "end": 11
+        }
+      ],
+      "note": "modes.md §3.9.3: a token between two holes sees opaque content on both sides."
+    },
+    {
+      "id": "de-de-protect-dash-adjacent-growth-declined",
+      "rule": "dashes",
+      "mode": "text",
+      "in": "${a}--${b}",
+      "protect": [
+        {
+          "start": 0,
+          "end": 4
+        },
+        {
+          "start": 6,
+          "end": 10
+        }
+      ],
+      "out": "${a}--${b}",
+      "protectOut": [
+        {
+          "start": 0,
+          "end": 4
+        },
+        {
+          "start": 6,
+          "end": 10
+        }
+      ],
+      "note": "modes.md §3.9.4, accepted cost: '--' → ' – ' grows by one code point at both edges of its span, so the edge-growth rule (§3.4) discards it. The same string with no protect converts (§7.9)."
+    },
+    {
+      "id": "de-de-protect-quotes-around-hole",
+      "rule": "quotes",
+      "mode": "text",
+      "in": "Er sagte \"${name}\" zweimal.",
+      "protect": [
+        {
+          "start": 10,
+          "end": 17
+        }
+      ],
+      "out": "Er sagte „${name}“ zweimal.",
+      "protectOut": [
+        {
+          "start": 10,
+          "end": 17
+        }
+      ],
+      "note": "modes.md §3.9.3: the same pairing in a locale whose glyphs differ in shape."
     }
   ]
 }

data/lib/polytypo/data/fixtures/el.json

@@ -1,5 +1,5 @@
 {
-  "spec": "1.12.0",
+  "spec": "1.14.0",
   "locale": "el",
   "cases": [
     {

data/lib/polytypo/data/fixtures/en-GB.json

@@ -1,5 +1,5 @@
 {
-  "spec": "1.12.0",
+  "spec": "1.14.0",
   "locale": "en-GB",
   "cases": [
     {
@@ -1388,6 +1388,94 @@
       "in": "(aside)'quoted' here",
       "out": "(aside)’quoted’ here",
       "note": "apostrophe.md §6 row 25 and §7 item 9 — case 2a's accepted cost, pinned so no port can narrow or widen it silently. The author meant a quotation and both marks now read as closing glyphs. It was mismatched before spec 1.5.0 too, in the other direction: 1.4.0 gave (aside)'quoted’ here, case 3 having curled the trailing mark while the leading one matched no case. `quotes` declines the leading mark because canOpen's right-test rejects a CLOSEISH left neighbour, so a quotation opening flush after a closing delimiter is exactly the shape it cannot pair, and two neighbours cannot separate that from a possessive."
+    },
+    {
+      "id": "en-gb-protect-leading-hole-dash",
+      "rule": "dashes",
+      "mode": "text",
+      "in": "${x} - Rules",
+      "protect": [
+        {
+          "start": 0,
+          "end": 4
+        }
+      ],
+      "out": "${x} – Rules",
+      "protectOut": [
+        {
+          "start": 0,
+          "end": 4
+        }
+      ],
+      "note": "modes.md §3.9.3: the same hole in a spaced-dash locale; the three-code-point token ' - ' becomes ' – ' without touching the hole."
+    },
+    {
+      "id": "en-gb-protect-trailing-hole-dash",
+      "rule": "dashes",
+      "mode": "text",
+      "in": "Rules - ${x}",
+      "protect": [
+        {
+          "start": 8,
+          "end": 12
+        }
+      ],
+      "out": "Rules – ${x}",
+      "protectOut": [
+        {
+          "start": 8,
+          "end": 12
+        }
+      ],
+      "note": "modes.md §3.9.3: a protected run at the end of the input is a marker too."
+    },
+    {
+      "id": "en-gb-protect-icu-quoted-brace",
+      "rule": "quotes",
+      "mode": "text",
+      "in": "Call it '{b}' now.",
+      "protect": [
+        {
+          "start": 8,
+          "end": 13
+        }
+      ],
+      "out": "Call it '{b}' now.",
+      "protectOut": [
+        {
+          "start": 8,
+          "end": 13
+        }
+      ],
+      "note": "polytypo/polytypo#14: in ICU MessageFormat an apostrophe before a brace is an escape. The caller protects the quoted brace including both apostrophes and they stay straight; without protect they become U+2018/U+2019."
+    },
+    {
+      "id": "en-gb-protect-touching-ranges",
+      "rule": "dashes",
+      "mode": "text",
+      "in": "a ${x}${y} - b",
+      "protect": [
+        {
+          "start": 2,
+          "end": 6
+        },
+        {
+          "start": 6,
+          "end": 10
+        }
+      ],
+      "out": "a ${x}${y} – b",
+      "protectOut": [
+        {
+          "start": 2,
+          "end": 6
+        },
+        {
+          "start": 6,
+          "end": 10
+        }
+      ],
+      "note": "modes.md §3.9.2: touching ranges are one protected run and give one marker."
     }
   ]
 }