/LPeg-Parsers

Permalink
Switch branches/tags
Nothing to show
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
Raw Blame History
493 lines (400 sloc) 10.3 KB
The code herein contains LPeg [1] routines for parsing some common data
formats. The current formats are:
abnf
The core ruleset from RFC-5234. These rules are used often in RFCs.
email
Parses email headers as defined in:
RFC-0822 Internet Message Format
RFC-1036 Standard for Interchange of USENET Messages
RFC-2045 Multipurpose Internet Mail Extensions I
RFC-2046 Multipurpose Internet Mail Extensions II
RFC-2047 Multipurpose Internet Mail Extensions III
RFC-2048 Multipurpose Internet Mail Extensions IV
RFC-2369 The Use of URLs as Meta-Syntax for Core Mail
List Commands and their Transport through
Message Header Fields
RFC-2822 Internet Message Format
RFC-2919 A Structured Field and Namespace for the Identification of Mailing Lists
RFC-5064 The Archived-At Message Header Field
RFC-5322 Internet Message Format
Headers are returned in a Lua table. For example, the following
headers:
Return-Path: <sean@conman.org>
Received: from brevard.conman.org (brevard.conman.org
[66.252.224.242])
by mail.example.com (Postfix)
with ESMTP id 538562EA5D07
for <sherlock@example.com>;
Fri, 28 Dec 2012 21:40:11 -0500
From: Sean Conner <sean@conman.org>
To: Sherlock Holmes <sherlock@example.com>,
the-scooby-gang: Fred <fred@example.net>,
Daphne <daphne@example.net>,
Velma <velma@example.net>,
Shaggy <shaggy@example.net>,
Scobby-Doo <scooby@example.net>;,
The Batman <batman@example.org>
Subject: I know who did it!
Date: Fri, 28 Dec 2012 21:40:11 -0500
Message-ID: <1234.5678.90abcd@conman.org>
Will return the following Lua table:
{
received =
{
[1] =
{
with = "ESMTP",
from = "brevard.conman.org",
id = "538562EA5D07",
when =
{
min = 0.000000,
zone = -18000.000000,
day = 28.000000,
month = 12.000000,
year = 2012.000000,
sec = 1.000000,
hour = 1.000000,
weekday = "Fri",
},
for =
{
address = "sherlock@example.com",
},
by = "mail.example.com",
},
},
to =
{
[1] =
{
name = "Sherlock Holmes",
address = "sherlock@example.com",
},
[2] =
{
['the-scooby-gang'] =
{
[1] =
{
name = "Fred",
address = "fred@example.net",
},
[2] =
{
name = "Daphne",
address = "daphne@example.net",
},
[3] =
{
name = "Velma",
address = "velma@example.net",
},
[4] =
{
name = "Shaggy",
address = "shaggy@example.net",
},
[5] =
{
name = "Scobby-Doo",
address = "scooby@example.net",
},
},
},
[3] =
{
name = "The Batman",
address = "batman@example.org",
},
},
from =
{
[1] =
{
name = "Sean Conner",
address = "sean@conman.org",
},
},
date =
{
min = 0.000000,
zone = -18000.000000,
day = 28.000000,
month = 12.000000,
year = 2012.000000,
sec = 1.000000,
hour = 1.000000,
weekday = "Fri",
},
return_path =
{
[1] =
{
address = "sean@conman.org",
},
},
message_id = "1234.5678.90abcd@conman.org",
subject = "I know who did it!",
}
The only fields not supported are the Resent-* fields; they are
rarely used and the semantics are particularly hard to support via
parsing only. These fields, as well as any other fields not
otherwise understood or parsable will end up on a field called
'generic' with the key being the raw header name.
json
Implements a JSON parser. It requires some additional modules [2]
to run. This will parse a JSON file into a Lua table. The full
grammar is supported, but it expects the input to be valid UTF-8.
A JSON null value will be converted to nil. If you won't want this
behavior, define a global variable called "null" to be the value
you want for a JSON null.
jsons
Another implementation of a JSON parser. This one "streams" the
input, meaning it will handle large JSON files the other one won't,
and is a drop in replacement. You can also pass in a function that
will return more data so you can actually "stream" data into the
parser.
ip
Provides two LPeg patterns, IPv4 and IPv6 which parse and convert
said addresses directly into their network-order binary formats.
ip-text
Provides two LPeg patterns, IPv4 and IPv6 which parse and return said
addresses as text, unlike the ip module above.
ini
Provides a INI file parser that returns a Lua table from a INI
file. A sample INI file such as:
; we allow "default" values
default = ok
[section1]
var1 = foo
var2 = 12,23,34,54,44
VAR3 = "var3=foo",33,44,55
var2 = apple
Var4 = 55
[section2]
# another comment
; and so is this one
var1 = foo bar baz ; this is a comment
var2 = "foo bar baz ; this is not a comment"
[section1]
var4=this is a test
var5= this is also a test
var2 = pear
var3 = 88,99
will result in a Lua table of:
{
section1 =
{
var1 = "foo",
var5 = "this is also a test",
var4 =
{
[1] = "55",
[2] = "this is a test",
},
var3 =
{
[1] = "var3=foo",
[2] = "33",
[3] = "44",
[4] = "55",
[5] = "88",
[6] = "99",
},
var2 =
{
[1] = "12",
[2] = "23",
[3] = "34",
[4] = "54",
[5] = "44",
[6] = "apple",
[7] = "pear",
},
},
default = "ok",
section2 =
{
var1 = "foo bar baz ",
var2 = "foo bar baz ; this is not a comment",
},
}
strftime
Parses the format string for strftime() (or os.date() for Lua) and
returns an LPeg expression that can parse that format, with the
exceptions of "%c", "%x" and "%X" (all system specific formats).
For example, the format, "%A, %d %B %Y @ %H:%M:%S" will return the
LPeg expression to parse this:
Monday, 02 July 2018 @ 16:02:48
into
{
min = 2.000000,
wday = 2.000000,
day = 2.000000,
month = 7.000000,
sec = 48.000000,
hour = 16.000000,
year = 2018.000000,
}
This will even work with other locales, such as "se_NO.UTF-8",
which will allow LPeg to parse:
vuossárga, 02 suoidnemánu 2018 @ 16:02:48
into
{
min = 2,000000,
wday = 2,000000,
day = 2,000000,
month = 7,000000,
sec = 48,000000,
hour = 16,000000,
year = 2018,000000,
}
url.url
Parses URLs per RFC-3986. By default, it will handle the following
URL types:
http:
https:
file:
ftp:
Given the following URL:
http://www.conman.org/people/spc/index.cgi?one=1%3F&two=2&three=3#target1
It will be broken down into a Lua table as follows:
{
host = "www.conman.org",
path =
{
[1] = "people",
[2] = "spc",
[3] = "index.cgi",
root = true,
},
query =
{
one = "1?",
three = "3",
two = "2",
},
scheme = "http",
fragment = "target1",
port = 80.000000,
}
The path is broken down by default. No normalization is done, so a
path like:
/foo/../bar/baz/../../snafu/./fubar
will be returned as:
{
[1] = "foo",
[2] = "..",
[3] = "bar",
[4] = "baz",
[5] = "..",
[6] = "..",
[7] = "snafu",
[8] = ".",
[9] = "fubar"
}
Other URLs can be parsed, but a URL like:
mailto:sean@conman.org?cc=fred@example.com,velma@example.net&subject=Current%20Mystery
will be broken down as:
{
scheme = "mailto",
path =
{
[1] = "sean@conman.org",
},
query =
{
cc = "fred@example.com,velma@example.net;subject",
subject = "Current Mystery",
},
}
which may require more parsing than provided here.
url.gopher
Parses "gopher:" URLs per RFC-4266. Given this URL:
gopher://gopher.conman.org/0foobar%09search%20String%09plus
it will be broken down as:
{
host = "gopher.conman.org",
type = "0",
search = "search String",
scheme = "gopher",
plus = "plus",
port = 70.000000,
selector = "foobar",
}
If you need to parse other URLs in addition to "gopher:" types,
you can do:
gopher = require "org.conman.parsers.url.url"
url = require "org.conman.parsers.url.url"
url = gopher + url
info = url:match(my_url)
url.sip
Parses "sip:" URIs per RFC-3261. Examples:
sip = require "org.conman.parsers.url.sip"
u = sip:match [[sip:annc@example.com;play=file://fs.example.net//clips/my-intro.dvi;content-type=video/mpeg%3bencode%d3314M-25/625-50]]
results in:
{
host = "example.com",
port = 5060.000000,
user = "annc",
scheme = "sip",
parameters =
{
play = "file://fs.example.net//clips/my-intro.dvi",
["content-type"] = "video/mpeg%3bencode%d3314M-25/625-50",
},
}
and
u = sip:match [[sip:+1-(555)-555-1212;ext=1234@example.net;user=phone]]
results in:
{
host = "example.net",
port = 5060.000000,
user =
{
number = "15555551212",
global = true,
parameters =
{
ext = "1234",
},
},
scheme = "sip",
parameters =
{
user = "phone",
},
}
If you need to parse other URLs in addition to "sip:" types,
you can do:
sip = require "org.conman.parsers.url.sip"
url = require "org.conman.parsers.url.url"
url = sip + url
info = url:match(my_url)
url.tel
Parses "tel:" URIs per RFC-3966. Example:
tel = require "org.conman.parsers.url.tel"
u = tel:match "tel:+1-(555)-555-1212;ext=1234"
results in:
{
scheme = "tel",
number = "15555551212",
global = true,
parameters =
{
ext = "1234",
},
}
If you need to parse other URLs in addition to "tel:" types,
you can do:
tel = require "org.conman.parsers.url.tel"
url = require "org.conman.parsers.url.url"
url = tel.tel + url
info = url:match(my_url)
NOTE: Unlike the other modules, this one returns a table instead
of an LPeg expression, due to some other requirements.
soundex.lua
Implements the Soundex algorithm.
[1] http://www.inf.puc-rio.br/~roberto/lpeg/
[2] https://github.com/spc476/lua-conmanorg