From 90bd50866590f390061a849b870c9b47a6741fd8 Mon Sep 17 00:00:00 2001
From: niksedk <nikse.dk@gmail.com>
Date: Sun, 19 Jul 2026 14:48:19 +0200
Subject: [PATCH] OCR fix lists: adopt the safest community rules from #12644
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

fraternl shared user OCR fix lists for English, French, Dutch and
Spanish refined over years (#12644). Adopted only the rules that fix
unambiguous OCR errors and cannot fire on legitimate text - every rule
was behaviorally verified in .NET with fix cases and must-not-change
cases, and deduped against the shipped lists.

eng: iihrer->ührer (ü misread as ii), Sefior->Señor family (ñ as fi),
compafiero/acompafiar. fra: ILy a, d'ou->d'où, 5'il->S'il, spaced
apostrophes (s 'il, t' aime), I'a->l'a, 5O5->SOS (letter O only - the
contributed rule also matched the number 505 and was narrowed), plus
dictionary-gated mère/père/très/Là diacritic fixes in the
IfSpelledCorrectly section. spa: iihrer, niñal->niña! and túl->tú!
(! misread as l). nld: doubled-apostrophe ''n/''t collapse,
wanner->wanneer.

Deliberately NOT adopted: style rules (imperative "Stop." -> "Stop!",
'n->een rewrites, music-symbol removal, ellipsis/a.m. normalization),
rules that corrupt legitimate text ("he cause"->because matches "did he
cause", solucionario is a real word, Neel is a name, "gebeurd hier" can
be valid word order, EI is real Spanish usage, 505 room numbers,
accepted spellings like facade/voila/deja vu), and rules with bugs in
the contributed files (doubled ${3}, $[ in replacement).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
---
 Dictionaries/eng_OCRFixReplaceList.xml |  6 ++++++
 Dictionaries/fra_OCRFixReplaceList.xml | 23 ++++++++++++++++++++++-
 Dictionaries/nld_OCRFixReplaceList.xml |  4 ++++
 Dictionaries/spa_OCRFixReplaceList.xml |  5 +++++
 4 files changed, 37 insertions(+), 1 deletion(-)

diff --git a/Dictionaries/eng_OCRFixReplaceList.xml b/Dictionaries/eng_OCRFixReplaceList.xml
index 1636eeed627..da82ad94aa4 100644
--- a/Dictionaries/eng_OCRFixReplaceList.xml
+++ b/Dictionaries/eng_OCRFixReplaceList.xml
@@ -3395,6 +3395,12 @@
 		<RegEx find="(2 )(nd\b)" replaceWith="2$2" />
 		<RegEx find="(3 )(rd\b)" replaceWith="3$2" />
 		<RegEx find="([456789])( th\b)" replaceWith="$1th" />
+		<!-- ü misread as ii: Fiihrer/Obergruppenfiihrer etc. (from #12644) -->
+		<RegEx find="iihrer" replaceWith="ührer" />
+		<!-- ñ misread as fi: Sefior/sefiorita, compafiero (from #12644) -->
+		<RegEx find="\b([Ss])efior" replaceWith="${1}eñor" />
+		<RegEx find="ompafier" replaceWith="ompañer" />
+		<RegEx find="ompafiar" replaceWith="ompañar" />
 	</RegularExpressions>
 	<RegularExpressionsIfSpelledCorrectly>
 		<RegEx find="\[([A-Zl]+)\]" replaceAllFrom="l" replaceAllTo="I" spellCheck="$1" replaceWith="[$1]" />		
diff --git a/Dictionaries/fra_OCRFixReplaceList.xml b/Dictionaries/fra_OCRFixReplaceList.xml
index 9e73675d20e..17bcbf0bd6a 100644
--- a/Dictionaries/fra_OCRFixReplaceList.xml
+++ b/Dictionaries/fra_OCRFixReplaceList.xml
@@ -267,7 +267,23 @@
   </PartialLines>
   <BeginLines />
   <EndLines />
-  <RegularExpressions />
+  <RegularExpressions>
+    <!-- From #12644 (community-contributed, safety-reviewed) -->
+    <!-- "Il y a" misread as "ILy a" -->
+    <RegEx find="\bILy a\b" replaceWith="Il y a" />
+    <!-- d'ou is always d'où -->
+    <RegEx find="\b([Dd])'([Oo])u\b" replaceWith="${1}'${2}ù" />
+    <!-- S misread as 5 in S'il -->
+    <RegEx find="5 *'il\b" replaceWith="S'il" />
+    <!-- spaced apostrophe: s 'il / s 'ils -->
+    <RegEx find="([Ss]) 'il(s?)\b" replaceWith="${1}'il${2}" />
+    <!-- l misread as I in l'a -->
+    <RegEx find="\bI'a\b" replaceWith="l'a" />
+    <!-- spaced apostrophe after t' before vowel/h: "Je t' aime" -->
+    <RegEx find=" t' ([aehiouyéèêàùâîôûAEHIOUYÉÈÊÀÙÂÎÔÛ])" replaceWith=" t'${1}" />
+    <!-- S misread as 5 in SOS (letter O only - never the number 505) -->
+    <RegEx find="\b5(\.*)O(\.*)5(\.*)\b" replaceWith="S${1}O${2}S${3}" />
+  </RegularExpressions>
   <RegularExpressionsIfSpelledCorrectly>
 
     <!-- ============================================================ -->
@@ -372,5 +388,10 @@
     <!-- inLondon → in London -->
     <RegEx find="\bin([A-Z][a-zæøåöäéèàùâêîôûëïœç]+)\b" spellCheck="$1" replaceWith="in $1" />
 
+
+    <!-- From #12644: è/à misread - only applied when the fixed word is in the dictionary -->
+    <RegEx find="\b([MmPp])[eéêë]re\b" spellCheck="${1}ère" replaceWith="${1}ère" />
+    <RegEx find="\b([Tt])r[eéêë]s\b" spellCheck="${1}rès" replaceWith="${1}rès" />
+    <RegEx find="\bL[äáâ]\b" spellCheck="Là" replaceWith="Là" />
   </RegularExpressionsIfSpelledCorrectly>
 </OCRFixReplaceList>
\ No newline at end of file
diff --git a/Dictionaries/nld_OCRFixReplaceList.xml b/Dictionaries/nld_OCRFixReplaceList.xml
index 5f3f1cbc626..3c5668754e1 100644
--- a/Dictionaries/nld_OCRFixReplaceList.xml
+++ b/Dictionaries/nld_OCRFixReplaceList.xml
@@ -108,6 +108,8 @@
     <Word from="ziin" to="zijn" />
     <Word from="zonderjou" to="zonder jou" />
     <Word from="Zonderjou" to="Zonder jou" />
+    <!-- from #12644 -->
+    <Word from="wanner" to="wanneer" />
   </WholeWords>
   <PartialWordsAlways />
   <PartialWords>
@@ -137,6 +139,8 @@
     <RegEx find="\beIk" replaceWith="elk" />
     <RegEx find="\bler(land|se|s|)\b" replaceWith="Ier$1" />
     <RegEx find="comite\b" replaceWith="comité" />
+    <!-- doubled apostrophe before 'n / 't (from #12644) -->
+    <RegEx find="''([nt])\b" replaceWith="'${1}" />
   </RegularExpressions>
   <RegularExpressionsIfSpelledCorrectly>
 
diff --git a/Dictionaries/spa_OCRFixReplaceList.xml b/Dictionaries/spa_OCRFixReplaceList.xml
index 5078478508f..1d10b4d9ac3 100644
--- a/Dictionaries/spa_OCRFixReplaceList.xml
+++ b/Dictionaries/spa_OCRFixReplaceList.xml
@@ -366,6 +366,9 @@
     <Word from="Bartomeu" to="Bertomeu" />
     <Word from="Iocalizarlo" to="localizarlo" />
     <Word from="Ilámame" to="llámame" />
+    <!-- ! misread as l (from #12644) -->
+    <Word from="niñal" to="niña!" />
+    <Word from="túl" to="tú!" />
   </WholeWords>
   <PartialWordsAlways />
   <PartialWords />
@@ -948,5 +951,7 @@
     <RegEx find="\Bi(log[ao]s?\b)" replaceWith="í$1" />
     <RegEx find="\bIes\b" replaceWith="les" />
     <RegEx find="\bIos\b" replaceWith="los" />
+    <!-- ü misread as ii: Fiihrer/Obergruppenfiihrer etc. (from #12644) -->
+    <RegEx find="iihrer" replaceWith="ührer" />
   </RegularExpressions>
 </OCRFixReplaceList>
\ No newline at end of file