runtime(python): highlight bytes in python

- Highlight bytes literals
- Do not highlight Unicode escape sequences in bytes literals

fixes: #14033
fixes: #17726
closes: #17728

Signed-off-by: Rob B <github@0x7e.net>
Signed-off-by: Christian Brabandt <cb@256bit.org>

diff --git a/runtime/syntax/python.vim b/runtime/syntax/python.vim
index 68036f5905bf26f6c6ffa26e0f0ff4c6409ded5c..759af2d4ffc36f9056583ecd9dd636e3b386f934 100644 (file)
--- a/runtime/syntax/python.vim
+++ b/runtime/syntax/python.vim
@@ -1,9 +1,10 @@
 " Vim syntax file
 " Language:    Python
 " Maintainer:  Zvezdan Petkovic <zpetkovic@acm.org>
-" Last Change: 2023 Feb 28
+" Last Change: 2025 Jul 14
 " Credits:     Neil Schemenauer <nas@python.ca>
 "              Dmitry Vasiliev
+"              Rob B
 "
 "              This version is a major rewrite by Zvezdan Petkovic.
 "
@@ -144,24 +145,48 @@ syn keyword pythonTodo            FIXME NOTE NOTES TODO XXX contained
 " Triple-quoted strings can contain doctests.
 syn region  pythonString matchgroup=pythonQuotes
       \ start=+[uU]\=\z(['"]\)+ end="\z1" skip="\\\\\|\\\z1"
-      \ contains=pythonEscape,@Spell
+      \ contains=pythonEscape,pythonUnicodeEscape,@Spell
 syn region  pythonString matchgroup=pythonTripleQuotes
       \ start=+[uU]\=\z('''\|"""\)+ end="\z1" keepend
-      \ contains=pythonEscape,pythonSpaceError,pythonDoctest,@Spell
+      \ contains=pythonEscape,pythonUnicodeEscape,pythonSpaceError,pythonDoctest,@Spell
 syn region  pythonRawString matchgroup=pythonQuotes
-      \ start=+[uU]\=[rR]\z(['"]\)+ end="\z1" skip="\\\\\|\\\z1"
+      \ start=+[rR]\z(['"]\)+ end="\z1" skip="\\\\\|\\\z1"
       \ contains=@Spell
 syn region  pythonRawString matchgroup=pythonTripleQuotes
-      \ start=+[uU]\=[rR]\z('''\|"""\)+ end="\z1" keepend
+      \ start=+[rR]\z('''\|"""\)+ end="\z1" keepend
       \ contains=pythonSpaceError,pythonDoctest,@Spell
 
+" Bytes
+syn region  pythonBytes
+      \ matchgroup=pythonQuotes
+      \ start=+\cB\z(['"]\)+
+      \ end="\z1"
+      \ skip="\\\\\|\\\z1"
+      \ contains=pythonEscape
+syn region  pythonBytes
+      \ matchgroup=pythonTripleQuotes
+      \ start=+\cB\z('''\|"""\)+
+      \ end="\z1"
+      \ keepend
+      \ contains=pythonEscape
+syn region  pythonRawBytes
+      \ matchgroup=pythonQuotes
+      \ start=+\c\%(BR\|RB\)\z(['"]\)+
+      \ end="\z1"
+      \ skip="\\\\\|\\\z1"
+syn region  pythonRawBytes
+      \ matchgroup=pythonTripleQuotes
+      \ start=+\c\%(BR\|RB\)\z('''\|"""\)+
+      \ end="\z1"
+      \ keepend
+
 syn match   pythonEscape       +\\[abfnrtv'"\\]+ contained
 syn match   pythonEscape       "\\\o\{1,3}" contained
 syn match   pythonEscape       "\\x\x\{2}" contained
-syn match   pythonEscape       "\%(\\u\x\{4}\|\\U\x\{8}\)" contained
+syn match   pythonUnicodeEscape        "\%(\\u\x\{4}\|\\U\x\{8}\)" contained
 " Python allows case-insensitive Unicode IDs: http://www.unicode.org/charts/
 " The specification: https://www.unicode.org/versions/Unicode16.0.0/core-spec/chapter-4/#G135165
-syn match   pythonEscape       "\\N{\a\+\%(\%(\s\a\+[[:alnum:]]*\)\|\%(-[[:alnum:]]\+\)\)*}" contained
+syn match   pythonUnicodeEscape        "\\N{\a\+\%(\%(\s\a\+[[:alnum:]]*\)\|\%(-[[:alnum:]]\+\)\)*}" contained
 syn match   pythonEscape       "\\$"
 
 " It is very important to understand all details before changing the
@@ -313,9 +338,12 @@ hi def link pythonComment          Comment
 hi def link pythonTodo                 Todo
 hi def link pythonString               String
 hi def link pythonRawString            String
+hi def link pythonBytes                String
+hi def link pythonRawBytes             String
 hi def link pythonQuotes               String
 hi def link pythonTripleQuotes         pythonQuotes
 hi def link pythonEscape               Special
+hi def link pythonUnicodeEscape                pythonEscape
 if !exists("python_no_number_highlight")
   hi def link pythonNumber             Number
 endif

diff --git a/runtime/syntax/testdir/dumps/python_strings_bytes_00.dump b/runtime/syntax/testdir/dumps/python_strings_bytes_00.dump
new file mode 100644 (file)
index 0000000..278acab
--- /dev/null
+++ b/runtime/syntax/testdir/dumps/python_strings_bytes_00.dump
@@ -0,0 +1,20 @@
+>#+0#0000e05#ffffff0| |S|t|r|i|n|g| |a|n|d| |B|y|t|e|s| |l|i|t|e|r|a|l|s| +0#0000000&@47
+|#+0#0000e05&| |h|t@1|p|s|:|/@1|d|o|c|s|.|p|y|t|h|o|n|.|o|r|g|/|3|/|r|e|f|e|r|e|n|c|e|/|l|e|x|i|c|a|l|_|a|n|a|l|y|s|i|s|.|h|t|m|l|#|s|t|r|i|n|g|-|a|n|d|-|b|y|t|e
+|s|-|l|i|t|e|r|a|l|s| +0#0000000&@64
+@75
+|#+0#0000e05&| |S|t|r|i|n|g|s| +0#0000000&@65
+|t|e|s|t| |=| |'+0#e000002&|S|t|r|i|n|g| |w|i|t|h| |e|s|c|a|p|e|s| |\+0#e000e06&|'| +0#e000002&|a|n|d| |\+0#e000e06&|"| +0#e000002&|a|n|d| |\+0#e000e06&|t|'+0#e000002&| +0#0000000&@29
+|t|e|s|t| |=| |"+0#e000002&|S|t|r|i|n|g| |w|i|t|h| |e|s|c|a|p|e|s| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1|"+0#e000002&| +0#0000000&@32
+|t|e|s|t| |=| |'+0#e000002&|S|t|r|i|n|g| |w|i|t|h| |e|s|c|a|p|e|s| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5| +0#e000002&|a|n|d| |\+0#e000e06&|N|{|I|N|V|E|R|T|E|D| |E|X|C|L|A|M|A|T|I
+|O|N| |M|A|R|K|}|'+0#e000002&| +0#0000000&@65
+|t|e|s|t| |=| |"+0#e000002&|S|t|r|i|n|g| |w|i|t|h| |e|s|c|a|p|e|d| |\+0#e000e06&@1| +0#e000002&|b|a|c|k|s|l|a|s|h| |a|n|d| |i|g|n|o|r|e|d| |\+0#e000e06&| +0#0000000&@20
+|n+0#e000002&|e|w|l|i|n|e|"| +0#0000000&@66
+|t|e|s|t| |=| |'+0#e000002&@2|S|t|r|i|n|g| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@38
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|t| +0#e000002&|a|n|d| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5|'+0#e000002&@2| +0#0000000&@38
+|t|e|s|t| |=| |"+0#e000002&@2|S|t|r|i|n|g| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@38
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|t| +0#e000002&|a|n|d| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5|"+0#e000002&@2| +0#0000000&@38
+@75
+|#+0#0000e05&| |R|a|w| |s|t|r|i|n|g|s| +0#0000000&@61
+@57|1|,|1| @10|T|o|p| 

diff --git a/runtime/syntax/testdir/dumps/python_strings_bytes_01.dump b/runtime/syntax/testdir/dumps/python_strings_bytes_01.dump
new file mode 100644 (file)
index 0000000..2684b01
--- /dev/null
+++ b/runtime/syntax/testdir/dumps/python_strings_bytes_01.dump
@@ -0,0 +1,20 @@
+|a+0#e000002#ffffff0|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5|'+0#e000002&@2| +0#0000000&@38
+|t|e|s|t| |=| |"+0#e000002&@2|S|t|r|i|n|g| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@38
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|t| +0#e000002&|a|n|d| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5|"+0#e000002&@2| +0#0000000&@38
+@75
+>#+0#0000e05&| |R|a|w| |s|t|r|i|n|g|s| +0#0000000&@61
+|t|e|s|t| |=| |r+0#e000002&|'|R|a|w| |s|t|r|i|n|g| |w|i|t|h| |l|i|t|e|r|a|l| |\|'| |a|n|d| |\|"| |a|n|d| |\|t|'| +0#0000000&@24
+|t|e|s|t| |=| |R+0#e000002&|"|R|a|w| |s|t|r|i|n|g| |w|i|t|h| |l|i|t|e|r|a|l| |\|0|4|0| |a|n|d| |\|x|F@1|"| +0#0000000&@27
+|t|e|s|t| |=| |r+0#e000002&|'|R|a|w| |s|t|r|i|n|g| |w|i|t|h| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5| |a|n|d| |\|N|{|I|N|V|E|R|T|E|D| |E|X|C|L
+|A|M|A|T|I|O|N| |M|A|R|K|}|'| +0#0000000&@60
+|t|e|s|t| |=| |R+0#e000002&|"|R|a|w| |s|t|r|i|n|g| |w|i|t|h| |l|i|t|e|r|a|l| |\@1| |b|a|c|k|s|l|a|s|h|e|s| |a|n|d| |l|i|t|e|r|a|l| |\| +0#0000000&@13
+|n+0#e000002&|e|w|l|i|n|e|"| +0#0000000&@66
+|t|e|s|t| |=| |r+0#e000002&|'@2|R|a|w| |s|t|r|i|n|g| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@33
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|t| |a|n|d| |\|0|4|0| |a|n|d| |\|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5|'@2| +0#0000000&@38
+|t|e|s|t| |=| |R+0#e000002&|"@2|R|a|w| |s|t|r|i|n|g| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@33
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|t| |a|n|d| |\|0|4|0| |a|n|d| |\|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5|"@2| +0#0000000&@38
+@75
+@57|1|7|,|1| @9|1|5|%| 

diff --git a/runtime/syntax/testdir/dumps/python_strings_bytes_02.dump b/runtime/syntax/testdir/dumps/python_strings_bytes_02.dump
new file mode 100644 (file)
index 0000000..c804bc3
--- /dev/null
+++ b/runtime/syntax/testdir/dumps/python_strings_bytes_02.dump
@@ -0,0 +1,20 @@
+| +0&#ffffff0@74
+|#+0#0000e05&| |B|y|t|e|s| +0#0000000&@67
+|t|e|s|t| |=| |b+0#e000002&|'|B|y|t|e|s| |w|i|t|h| |e|s|c|a|p|e|s| |\+0#e000e06&|'| +0#e000002&|a|n|d| |\+0#e000e06&|"| +0#e000002&|a|n|d| |\+0#e000e06&|t|'+0#e000002&| +0#0000000&@29
+|t|e|s|t| |=| |B+0#e000002&|"|B|y|t|e|s| |w|i|t|h| |e|s|c|a|p|e|s| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1|"+0#e000002&| +0#0000000&@32
+|t|e|s|t| |=| |b+0#e000002&|'|B|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5| |a|n|d| |\|N|{|I|N|V|E|R|T|E|D| |E|X|C|L|A|M|A|T|I
+|O|N| |M|A|R|K|}|'| +0#0000000&@65
+>t|e|s|t| |=| |B+0#e000002&|"|B|y|t|e|s| |w|i|t|h| |e|s|c|a|p|e|d| |\+0#e000e06&@1| +0#e000002&|b|a|c|k|s|l|a|s|h| |a|n|d| |i|g|n|o|r|e|d| |\+0#e000e06&| +0#0000000&@20
+|n+0#e000002&|e|w|l|i|n|e|"| +0#0000000&@66
+|t|e|s|t| |=| |b+0#e000002&|'@2|B|y|t|e|s| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@38
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|t| +0#e000002&|a|n|d| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5|'@2| +0#0000000&@38
+|t|e|s|t| |=| |B+0#e000002&|"@2|B|y|t|e|s| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@38
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|t| +0#e000002&|a|n|d| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5|"@2| +0#0000000&@38
+@75
+|#+0#0000e05&| |R|a|w| |b|y|t|e|s| +0#0000000&@63
+|t|e|s|t| |=| |b+0#e000002&|r|'|R|a|w| |b|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\|'| |a|n|d| |\|"| |a|n|d| |\|t|'| +0#0000000&@24
+|t|e|s|t| |=| |b+0#e000002&|R|"|R|a|w| |b|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\|0|4|0| |a|n|d| |\|x|F@1|"| +0#0000000&@27
+|t|e|s|t| |=| |B+0#e000002&|r|'|R|a|w| |b|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5| |a|n|d| |\|N|{|I|N|V|E|R|T|E|D| |E|@+0#4040ff13&@2
+| +0#0000000&@56|3|4|,|1| @9|3|8|%| 

diff --git a/runtime/syntax/testdir/dumps/python_strings_bytes_03.dump b/runtime/syntax/testdir/dumps/python_strings_bytes_03.dump
new file mode 100644 (file)
index 0000000..27087d9
--- /dev/null
+++ b/runtime/syntax/testdir/dumps/python_strings_bytes_03.dump
@@ -0,0 +1,20 @@
+|t+0&#ffffff0|e|s|t| |=| |B+0#e000002&|r|'|R|a|w| |b|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5| |a|n|d| |\|N|{|I|N|V|E|R|T|E|D| |E|X|C|L
+|A|M|A|T|I|O|N| |M|A|R|K|}|'| +0#0000000&@60
+|t|e|s|t| |=| |B+0#e000002&|R|"|R|a|w| |b|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\@1| |b|a|c|k|s|l|a|s|h|e|s| |a|n|d| |l|i|t|e|r|a|l| |\| +0#0000000&@13
+|n+0#e000002&|e|w|l|i|n|e|"| +0#0000000&@66
+|t|e|s|t| |=| |r+0#e000002&|b|'|R|a|w| |b|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\|'| |a|n|d| |\|"| |a|n|d| |\|t|'| +0#0000000&@24
+>t|e|s|t| |=| |r+0#e000002&|B|"|R|a|w| |b|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\|0|4|0| |a|n|d| |\|x|F@1|"| +0#0000000&@27
+|t|e|s|t| |=| |R+0#e000002&|b|'|R|a|w| |b|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5| |a|n|d| |\|N|{|I|N|V|E|R|T|E|D| |E|X|C|L
+|A|M|A|T|I|O|N| |M|A|R|K|}|'| +0#0000000&@60
+|t|e|s|t| |=| |R+0#e000002&|B|"|R|a|w| |b|y|t|e|s| |w|i|t|h| |l|i|t|e|r|a|l| |\@1| |b|a|c|k|s|l|a|s|h|e|s| |a|n|d| |l|i|t|e|r|a|l| |\| +0#0000000&@13
+|n+0#e000002&|e|w|l|i|n|e|"| +0#0000000&@66
+|t|e|s|t| |=| |b+0#e000002&|r|'@2|R|a|w| |b|y|t|e|s| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@33
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|t| |a|n|d| |\|0|4|0| |a|n|d| |\|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5|'@2| +0#0000000&@38
+|t|e|s|t| |=| |R+0#e000002&|B|"@2|R|a|w| |b|y|t|e|s| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@33
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|t| |a|n|d| |\|0|4|0| |a|n|d| |\|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |l|i|t|e|r|a|l| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5|"@2| +0#0000000&@38
+@75
+|#+0#0000e05&| |U|n|i|c|o|d|e| |l|i|t|e|r|a|l|s|:| |P|r|e|f|i|x| |i|s| |a|l@1|o|w|e|d| |b|u|t| |i|g|n|o|r|e|d| |(|h|t@1|p|s|:|/@1|p|e|p|s|.|p|y|t|h|o|n|.|o|r|g|/
+|p|e|p|-|0|4|1|4|)| +0#0000000&@65
+@57|5|0|,|1| @9|6|0|%| 

diff --git a/runtime/syntax/testdir/dumps/python_strings_bytes_04.dump b/runtime/syntax/testdir/dumps/python_strings_bytes_04.dump
new file mode 100644 (file)
index 0000000..545e722
--- /dev/null
+++ b/runtime/syntax/testdir/dumps/python_strings_bytes_04.dump
@@ -0,0 +1,20 @@
+|#+0#0000e05#ffffff0| |U|n|i|c|o|d|e| |l|i|t|e|r|a|l|s|:| |P|r|e|f|i|x| |i|s| |a|l@1|o|w|e|d| |b|u|t| |i|g|n|o|r|e|d| |(|h|t@1|p|s|:|/@1|p|e|p|s|.|p|y|t|h|o|n|.|o|r|g|/
+|p|e|p|-|0|4|1|4|)| +0#0000000&@65
+|t|e|s|t| |=| |u+0#e000002&|'|S|t|r|i|n|g| |w|i|t|h| |e|s|c|a|p|e|s| |\+0#e000e06&|'| +0#e000002&|a|n|d| |\+0#e000e06&|"| +0#e000002&|a|n|d| |\+0#e000e06&|t|'+0#e000002&| +0#0000000&@28
+|t|e|s|t| |=| |U+0#e000002&|"|S|t|r|i|n|g| |w|i|t|h| |e|s|c|a|p|e|s| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1|"+0#e000002&| +0#0000000&@31
+|t|e|s|t| |=| |u+0#e000002&|'|S|t|r|i|n|g| |w|i|t|h| |e|s|c|a|p|e|s| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5| +0#e000002&|a|n|d| |\+0#e000e06&|N|{|I|N|V|E|R|T|E|D| |E|X|C|L|A|M|A|T
+|I|O|N| |M|A|R|K|}|'+0#e000002&| +0#0000000&@64
+>t|e|s|t| |=| |U+0#e000002&|"|S|t|r|i|n|g| |w|i|t|h| |e|s|c|a|p|e|d| |\+0#e000e06&@1| +0#e000002&|b|a|c|k|s|l|a|s|h| |a|n|d| |i|g|n|o|r|e|d| |\+0#e000e06&| +0#0000000&@19
+|n+0#e000002&|e|w|l|i|n|e|"| +0#0000000&@66
+|t|e|s|t| |=| |u+0#e000002&|'@2|S|t|r|i|n|g| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@37
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|t| +0#e000002&|a|n|d| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5|'+0#e000002&@2| +0#0000000&@38
+|t|e|s|t| |=| |U+0#e000002&|"@2|S|t|r|i|n|g| |w|i|t|h| |q|u|o|t|e|s| |'| |a|n|d| |"| +0#0000000&@37
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|t| +0#e000002&|a|n|d| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1| +0#0000000&@42
+|a+0#e000002&|n|d| |e|s|c|a|p|e|s| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5|"+0#e000002&@2| +0#0000000&@38
+@75
+|#+0#0000e05&| |R|a|w| |U|n|i|c|o|d|e| |l|i|t|e|r|a|l|s| |a|r|e| |n|o|t| |a|l@1|o|w|e|d| +0#0000000&@36
+|t|e|s|t| |=| |u|r+0#e000002&|'|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |\|'| |a|n|d| |\|"| |a|n|d| |\|t|'| +0#0000000&@27
+|t|e|s|t| |=| |u|R+0#e000002&|"|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |\|0|4|0| |a|n|d| |\|x|F@1|"| +0#0000000&@30
+|t|e|s|t| |=| |U|r+0#e000002&|'|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5| |a|n|d| |\|N|{|I|N|V|E|R|T|E|D| |E|X|C|L|@+0#4040ff13&@2
+| +0#0000000&@56|6|5|,|1| @9|8|1|%| 

diff --git a/runtime/syntax/testdir/dumps/python_strings_bytes_05.dump b/runtime/syntax/testdir/dumps/python_strings_bytes_05.dump
new file mode 100644 (file)
index 0000000..e403cb2
--- /dev/null
+++ b/runtime/syntax/testdir/dumps/python_strings_bytes_05.dump
@@ -0,0 +1,20 @@
+|t+0&#ffffff0|e|s|t| |=| |U|r+0#e000002&|'|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5| |a|n|d| |\|N|{|I|N|V|E|R|T|E|D| |E|X|C|L|A|M|A
+|T|I|O|N| |M|A|R|K|}|'| +0#0000000&@63
+|t|e|s|t| |=| |U|R+0#e000002&|"|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |\@1| |b|a|c|k|s|l|a|s|h|e|s| |a|n|d| |l|i|t|e|r|a|l| |\| +0#0000000&@16
+|n+0#e000002&|e|w|l|i|n|e|"| +0#0000000&@66
+|t|e|s|t| |=| |r|u+0#e000002&|'|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |\+0#e000e06&|'| +0#e000002&|a|n|d| |\+0#e000e06&|"| +0#e000002&|a|n|d| |\+0#e000e06&|t|'+0#e000002&| +0#0000000&@27
+>t|e|s|t| |=| |r|U+0#e000002&|"|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1|"+0#e000002&| +0#0000000&@30
+|t|e|s|t| |=| |R|u+0#e000002&|'|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5| +0#e000002&|a|n|d| |\+0#e000e06&|N|{|I|N|V|E|R|T|E|D| |E|X|C|L|A|M|A
+|T|I|O|N| |M|A|R|K|}|'+0#e000002&| +0#0000000&@63
+|t|e|s|t| |=| |R|U+0#e000002&|"|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |\+0#e000e06&@1| +0#e000002&|b|a|c|k|s|l|a|s|h|e|s| |a|n|d| |l|i|t|e|r|a|l| |\+0#e000e06&| +0#0000000&@16
+|n+0#e000002&|e|w|l|i|n|e|"| +0#0000000&@66
+|t|e|s|t| |=| |u|r+0#e000002&|'@2|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |'| |a|n|d| |"| +0#0000000&@35
+|a+0#e000002&|n|d| |\|t| |a|n|d| |\|0|4|0| |a|n|d| |\|x|F@1| +0#0000000&@50
+|a+0#e000002&|n|d| |\|u|0@1|A|1| |a|n|d| |\|U|0@2|1|0|6|0|5|'@2| +0#0000000&@46
+|t|e|s|t| |=| |R|U+0#e000002&|"@2|I|n|v|a|l|i|d| |s|t|r|i|n|g| |w|i|t|h| |'| |a|n|d| |"| +0#0000000&@35
+|a+0#e000002&|n|d| |\+0#e000e06&|t| +0#e000002&|a|n|d| |\+0#e000e06&|0|4|0| +0#e000002&|a|n|d| |\+0#e000e06&|x|F@1| +0#0000000&@50
+|a+0#e000002&|n|d| |\+0#e000e06&|u|0@1|A|1| +0#e000002&|a|n|d| |\+0#e000e06&|U|0@2|1|0|6|0|5|"+0#e000002&@2| +0#0000000&@46
+|~+0#4040ff13&| @73
+|~| @73
+|~| @73
+| +0#0000000&@56|8|1|,|1| @9|B|o|t| 

diff --git a/runtime/syntax/testdir/input/python_strings_bytes.py b/runtime/syntax/testdir/input/python_strings_bytes.py
new file mode 100644 (file)
index 0000000..b4a536c
--- /dev/null
+++ b/runtime/syntax/testdir/input/python_strings_bytes.py
@@ -0,0 +1,90 @@
+# String and Bytes literals
+# https://docs.python.org/3/reference/lexical_analysis.html#string-and-bytes-literals
+
+# Strings
+test = 'String with escapes \' and \" and \t'
+test = "String with escapes \040 and \xFF"
+test = 'String with escapes \u00A1 and \U00010605 and \N{INVERTED EXCLAMATION MARK}'
+test = "String with escaped \\ backslash and ignored \
+newline"
+test = '''String with quotes ' and "
+and escapes \t and \040 and \xFF
+and escapes \u00A1 and \U00010605'''
+test = """String with quotes ' and "
+and escapes \t and \040 and \xFF
+and escapes \u00A1 and \U00010605"""
+
+# Raw strings
+test = r'Raw string with literal \' and \" and \t'
+test = R"Raw string with literal \040 and \xFF"
+test = r'Raw string with literal \u00A1 and \U00010605 and \N{INVERTED EXCLAMATION MARK}'
+test = R"Raw string with literal \\ backslashes and literal \
+newline"
+test = r'''Raw string with quotes ' and "
+and literal \t and \040 and \xFF
+and literal \u00A1 and \U00010605'''
+test = R"""Raw string with quotes ' and "
+and literal \t and \040 and \xFF
+and literal \u00A1 and \U00010605"""
+
+# Bytes
+test = b'Bytes with escapes \' and \" and \t'
+test = B"Bytes with escapes \040 and \xFF"
+test = b'Bytes with literal \u00A1 and \U00010605 and \N{INVERTED EXCLAMATION MARK}'
+test = B"Bytes with escaped \\ backslash and ignored \
+newline"
+test = b'''Bytes with quotes ' and "
+and escapes \t and \040 and \xFF
+and literal \u00A1 and \U00010605'''
+test = B"""Bytes with quotes ' and "
+and escapes \t and \040 and \xFF
+and literal \u00A1 and \U00010605"""
+
+# Raw bytes
+test = br'Raw bytes with literal \' and \" and \t'
+test = bR"Raw bytes with literal \040 and \xFF"
+test = Br'Raw bytes with literal \u00A1 and \U00010605 and \N{INVERTED EXCLAMATION MARK}'
+test = BR"Raw bytes with literal \\ backslashes and literal \
+newline"
+test = rb'Raw bytes with literal \' and \" and \t'
+test = rB"Raw bytes with literal \040 and \xFF"
+test = Rb'Raw bytes with literal \u00A1 and \U00010605 and \N{INVERTED EXCLAMATION MARK}'
+test = RB"Raw bytes with literal \\ backslashes and literal \
+newline"
+test = br'''Raw bytes with quotes ' and "
+and literal \t and \040 and \xFF
+and literal \u00A1 and \U00010605'''
+test = RB"""Raw bytes with quotes ' and "
+and literal \t and \040 and \xFF
+and literal \u00A1 and \U00010605"""
+
+# Unicode literals: Prefix is allowed but ignored (https://peps.python.org/pep-0414)
+test = u'String with escapes \' and \" and \t'
+test = U"String with escapes \040 and \xFF"
+test = u'String with escapes \u00A1 and \U00010605 and \N{INVERTED EXCLAMATION MARK}'
+test = U"String with escaped \\ backslash and ignored \
+newline"
+test = u'''String with quotes ' and "
+and escapes \t and \040 and \xFF
+and escapes \u00A1 and \U00010605'''
+test = U"""String with quotes ' and "
+and escapes \t and \040 and \xFF
+and escapes \u00A1 and \U00010605"""
+
+# Raw Unicode literals are not allowed
+test = ur'Invalid string with \' and \" and \t'
+test = uR"Invalid string with \040 and \xFF"
+test = Ur'Invalid string with \u00A1 and \U00010605 and \N{INVERTED EXCLAMATION MARK}'
+test = UR"Invalid string with \\ backslashes and literal \
+newline"
+test = ru'Invalid string with \' and \" and \t'
+test = rU"Invalid string with \040 and \xFF"
+test = Ru'Invalid string with \u00A1 and \U00010605 and \N{INVERTED EXCLAMATION MARK}'
+test = RU"Invalid string with \\ backslashes and literal \
+newline"
+test = ur'''Invalid string with ' and "
+and \t and \040 and \xFF
+and \u00A1 and \U00010605'''
+test = RU"""Invalid string with ' and "
+and \t and \040 and \xFF
+and \u00A1 and \U00010605"""