Add Sieve lexer (#1257)

Add Sieve lexer

4 files changed, 126 insertions, 1 deletions

diff --git a/AUTHORS b/AUTHORS
index 58968f50..9b6083e3 100644
--- a/AUTHORS
+++ b/AUTHORS
@@ -32,7 +32,7 @@ Other contributors, listed alphabetically, are:
 * Sébastien Bigaret -- QVT Operational lexer
 * Jarrett Billingsley -- MiniD lexer
 * Adam Blinkinsop -- Haskell, Redcode lexers
-* Stéphane Blondon -- SGF lexer
+* Stéphane Blondon -- SGF and Sieve lexers
 * Frits van Bommel -- assembler lexers
 * Pierre Bourdon -- bugfixes
 * Martijn Braam -- Kernel log lexer
diff --git a/pygments/lexers/_mapping.py b/pygments/lexers/_mapping.py
index 1afd5df7..2ae0c375 100644
--- a/pygments/lexers/_mapping.py
+++ b/pygments/lexers/_mapping.py
@@ -398,6 +398,7 @@ LEXERS = {
     'ScssLexer': ('pygments.lexers.css', 'SCSS', ('scss',), ('*.scss',), ('text/x-scss',)),
     'ShExCLexer': ('pygments.lexers.rdf', 'ShExC', ('shexc', 'shex'), ('*.shex',), ('text/shex',)),
     'ShenLexer': ('pygments.lexers.lisp', 'Shen', ('shen',), ('*.shen',), ('text/x-shen', 'application/x-shen')),
+    'SieveLexer': ('pygments.lexers.sieve', 'Sieve', ('sieve',), ('*.siv', '*.sieve'), ()),
     'SilverLexer': ('pygments.lexers.verification', 'Silver', ('silver',), ('*.sil', '*.vpr'), ()),
     'SlashLexer': ('pygments.lexers.slash', 'Slash', ('slash',), ('*.sl',), ()),
     'SlimLexer': ('pygments.lexers.webmisc', 'Slim', ('slim',), ('*.slim',), ('text/x-slim',)),
diff --git a/pygments/lexers/sieve.py b/pygments/lexers/sieve.py
new file mode 100644
index 00000000..9d7dba35
--- /dev/null
+++ b/pygments/lexers/sieve.py
@@ -0,0 +1,68 @@
+# -*- coding: utf-8 -*-
+"""
+    pygments.lexers.sieve
+    ~~~~~~~~~~~~~~~~~~~~~
+
+    Lexer for Sieve file format.
+
+    https://tools.ietf.org/html/rfc5228
+    https://tools.ietf.org/html/rfc5173
+    https://tools.ietf.org/html/rfc5229
+    https://tools.ietf.org/html/rfc5230
+    https://tools.ietf.org/html/rfc5232
+    https://tools.ietf.org/html/rfc5235
+    https://tools.ietf.org/html/rfc5429
+    https://tools.ietf.org/html/rfc8580
+
+    :copyright: Copyright 2006-2019 by the Pygments team, see AUTHORS.
+    :license: BSD, see LICENSE for details.
+"""
+
+from pygments.lexer import RegexLexer, bygroups
+from pygments.token import Comment, Name, Literal, String, Text, Punctuation, Keyword
+
+__all__ = ["SieveLexer"]
+
+
+class SieveLexer(RegexLexer):
+    """
+    Lexer for sieve format.
+    """
+    name = 'Sieve'
+    filenames = ['*.siv', '*.sieve']
+
+    tokens = {
+        'root': [
+            (r'\s+', Text),
+            (r'[();,{}\[\]]', Punctuation),
+            # import:
+            (r'(?i)require',
+             Keyword.Namespace),
+            # tags:
+            (r'(?i)(:)(addresses|all|contains|content|create|copy|comparator|count|days|detail|domain|fcc|flags|from|handle|importance|is|localpart|length|lowerfirst|lower|matches|message|mime|options|over|percent|quotewildcard|raw|regex|specialuse|subject|text|under|upperfirst|upper|value)',
+             bygroups(Name.Tag, Name.Tag)),
+            # tokens:
+            (r'(?i)(address|addflag|allof|anyof|body|discard|elsif|else|envelope|ereject|exists|false|fileinto|if|hasflag|header|keep|notify_method_capability|notify|not|redirect|reject|removeflag|setflag|size|spamtest|stop|string|true|vacation|virustest)',
+             Name.Builtin),
+            (r'(?i)set',
+             Keyword.Declaration),
+            # number:
+            (r'([0-9.]+)([kmgKMG])?',
+             bygroups(Literal.Number, Literal.Number)),
+            # comment:
+            (r'#.*$',
+             Comment.Single),
+            (r'/\*.*\*/',
+             Comment.Multiline),
+            # string:
+            (r'"[^"]*?"',
+             String),
+            # text block:
+            (r'text:',
+             Name.Tag, 'text'),
+        ],
+        'text': [
+            (r'[^.].*?\n', String),
+            (r'^\.', Punctuation, "#pop"),
+        ]
+    }
diff --git a/tests/examplefiles/example.sieve b/tests/examplefiles/example.sieve
new file mode 100644
index 00000000..8abed421
--- /dev/null
+++ b/tests/examplefiles/example.sieve
@@ -0,0 +1,56 @@
+# Based on examples included the RFC files.
+# It doesn't show cases for every token parsed by the lexer.
+
+require ["fileinto", "envelope"];
+
+if address :is "to" "demo@example.org" {
+  fileinto "demo-list";
+} elsif envelope :is "from" "owner@example.org" {
+  fileinto "lists.cipe";
+} elsif (exists "X-Cron-Env",
+          header :regex    ["subject"] [".* A",
+                                        ".* B"]) {
+  addflag "Cron";
+  stop;
+} else {
+  keep;
+}
+
+if body :text :contains "project schedule" {
+        fileinto "project/schedule";
+}
+
+require "vacation";
+if header :contains "subject" "lunch" {
+    vacation :handle "ran-away" "I'm out and can't meet for lunch";
+} else {
+    vacation :handle "ran-away" "I'm out";
+}
+
+
+if virustest :value "eq" :comparator "i;ascii-numeric" "0" {
+  fileinto "Unscanned";
+  /* Infected with high probability (value range in 1-5) */
+}
+
+if envelope :detail "to" "spam"{
+  fileinto "Spam";
+}
+
+if size :over 100k {
+  discard;
+}
+elsif address :DOMAIN :is ["From", "To"] "example.com"
+{
+  keep;
+}
+
+require ["reject"];
+
+if size :over 1M {
+  reject text:
+Your message is too big.  If you want to send me a big attachment,
+put it on a public web site and send me a URL.
+.
+;
+}