From ff9fee2e1a16e195d4f569330462690024a13b58 Mon Sep 17 00:00:00 2001
From: Alexander Bezzubov <bzz@apache.org>
Date: Wed, 9 Jan 2019 21:13:57 +0100
Subject: [PATCH] gen: fix regexp Or syntax

Signed-off-by: Alexander Bezzubov <bzz@apache.org>
---
 data/content.go                                 |  4 ++--
 internal/code-generator/generator/heuristics.go | 11 +++++++----
 2 files changed, 9 insertions(+), 6 deletions(-)

diff --git a/data/content.go b/data/content.go
index 66f955ae..c595204b 100644
--- a/data/content.go
+++ b/data/content.go
@@ -195,7 +195,7 @@ var ContentHeuristics = map[string]*Heuristics{
 		},
 		&OrRule{
 			&Languages{[]string{"C++"}},
-			regexp.MustCompile(`(?m)^\s*#\s*include <(cstdint|string|vector|map|list|array|bitset|queue|stack|forward_list|unordered_map|unordered_set|(i|o|io)stream)> | ^\s*template\s*< | ^[ \t]*try | ^[ \t]*catch\s*\( | ^[ \t]*(class|(using[ \t]+)?namespace)\s+\w+ | ^[ \t]*(private|public|protected):$ | std::\w+`),
+			regexp.MustCompile(`(?m)^\s*#\s*include <(cstdint|string|vector|map|list|array|bitset|queue|stack|forward_list|unordered_map|unordered_set|(i|o|io)stream)>|^\s*template\s*<|^[ \t]*try|^[ \t]*catch\s*\(|^[ \t]*(class|(using[ \t]+)?namespace)\s+\w+|^[ \t]*(private|public|protected):$|std::\w+`),
 		},
 	},
 	".hh": &Heuristics{
@@ -300,7 +300,7 @@ var ContentHeuristics = map[string]*Heuristics{
 	".md": &Heuristics{
 		&OrRule{
 			&Languages{[]string{"Markdown"}},
-			regexp.MustCompile(`(?m)(^[-A-Za-z0-9=#!\*\[|>])|<\/ | \A\z`),
+			regexp.MustCompile(`(?m)(^[-A-Za-z0-9=#!\*\[|>])|<\/|\A\z`),
 		},
 		&OrRule{
 			&Languages{[]string{"GCC Machine Description"}},
diff --git a/internal/code-generator/generator/heuristics.go b/internal/code-generator/generator/heuristics.go
index 3eaabce4..4e75a97f 100644
--- a/internal/code-generator/generator/heuristics.go
+++ b/internal/code-generator/generator/heuristics.go
@@ -10,6 +10,11 @@ import (
 	yaml "gopkg.in/yaml.v2"
 )
 
+const (
+	multilinePrefix = "(?m)"
+	orPipe          = "|"
+)
+
 // GenHeuristics generates language identification heuristics in Go.
 // It is of generator.File type.
 func GenHeuristics(fileToParse, _, outPath, tmplPath, tmplName, commit string) error {
@@ -67,14 +72,14 @@ func loadRule(namedPatterns map[string]StringArray, rule *Rule) *LanguagePattern
 		}
 		result = &LanguagePattern{"And", rule.Languages, "", subPatterns}
 	} else if len(rule.Pattern) != 0 { // OrPattern
-		conjunction := strings.Join(rule.Pattern, " | ")
+		conjunction := strings.Join(rule.Pattern, orPipe)
 		pattern := convertToValidRegexp(conjunction)
 		result = &LanguagePattern{"Or", rule.Languages, pattern, nil}
 	} else if rule.NegativePattern != "" { // NotPattern
 		pattern := convertToValidRegexp(rule.NegativePattern)
 		result = &LanguagePattern{"Not", rule.Languages, pattern, nil}
 	} else if rule.NamedPattern != "" { // Named OrPattern
-		conjunction := strings.Join(namedPatterns[rule.NamedPattern], " | ")
+		conjunction := strings.Join(namedPatterns[rule.NamedPattern], orPipe)
 		pattern := convertToValidRegexp(conjunction)
 		result = &LanguagePattern{"Or", rule.Languages, pattern, nil}
 	} else { // AlwaysPattern
@@ -164,8 +169,6 @@ func isUnsupportedRegexpSyntax(reg string) bool {
 		(strings.HasPrefix(reg, multilinePrefix+`/`) && strings.HasSuffix(reg, `/`))
 }
 
-const multilinePrefix = "(?m)"
-
 // convertToValidRegexp converts Ruby regexp syntaxt to RE2 equivalent.
 // Does not work with Ruby regexp literals.
 func convertToValidRegexp(rubyRegexp string) string {