3
README: library "Sexplib"
4
*************************
5
Copyright (C) 2012 Jane Street Holding, LLC (1)
6
=====================================================
18
------------------------------------------------------------------------
19
| CHANGES | History of code changes |
20
------------------------------------------------------------------------
21
| COPYRIGHT | Notes on copyright |
22
------------------------------------------------------------------------
23
| INSTALL | Short notes on compiling and |
24
| | installing the library |
25
------------------------------------------------------------------------
26
| LICENSE | "GNU LESSER GENERAL PUBLIC LICENSE" |
27
------------------------------------------------------------------------
28
| LICENSE.Tywith | License of Tywith, from which Sexplib is derived|
29
------------------------------------------------------------------------
30
| Makefile | Top Makefile |
31
------------------------------------------------------------------------
32
| OCamlMakefile | Generic Makefile for OCaml-projects |
33
------------------------------------------------------------------------
34
| OMakefile | Ignore this file |
35
------------------------------------------------------------------------
36
| README.txt | This file |
37
------------------------------------------------------------------------
38
| lib/ | OCaml-library for S-expression conversions |
39
------------------------------------------------------------------------
40
| lib_test/ | Test applications for the Sexplib-library |
41
------------------------------------------------------------------------
50
This library contains functionality for parsing and pretty-printing
51
S-expressions. In addition to that it contains a preprocessing module for
52
Camlp4, which can be used to automatically generate code from type definitions
53
for efficiently converting OCaml-values to S-expressions and vice versa. In
54
combination with the parsing and pretty-printing functionality this frees
55
users from having to write their own I/O-routines for datastructures they
56
define. Possible errors during automatic conversions from S-expressions to
57
OCaml-values are reported in human-readable ways with exact location
58
information. Another module in this library allows you to extract and replace
59
sub-expressions in S-expressions.
63
*=*=*=*=*=*=*=*=*=*=*=
66
Make sure you have installed the 'type_conv' package on your system, too. It
67
should be obtainable at the same site as 'sexplib'.
69
The API (.mli-files) in the 'sexplib' library directory is fully documented.
70
Module 'Sexp' contains all I/O-functions for S-expressions, module 'Conv'
71
helper functions for converting OCaml-values of standard types to
72
S-expressions. Module 'Path' supports sub-expression extraction and
75
Module 'pa_sexp_conv.ml' contains the extensions for the
76
Camlp4-preprocessor. It adds the new construct 'with sexp' (and 'with sexp_of'
77
and 'with of_sexp', which are implied by the first). When using this construct
78
right after a type definition, function definitions will be generated
79
automatically, which perform S-expression conversions.
81
E.g. given the following type definition:
86
The above will generate the functions 'sexp_of_t' and 't_of_sexp'. The
87
preprocessor also supports automatic addition of conversion functions to
88
signatures. Just add 'with sexp' to the type in a signature, and the
89
appropriate function signatures will be generated.
91
Converters for standard types (int, list, Hashtbl.t, etc.) become visible to
92
the macro-generated code by opening the standard module before their first use
93
in a type definition. Users will therefore usually want to place the following
94
at the top of their files:
98
See the file 'lib_test/conv_test.ml' for example usage. It also demonstrates
99
how to extract and substitute sub-expressions.
102
3.1 Compiling and linking
103
==========================
105
To compile a file you will have to add a preprocessing flag to the compiler
106
invocation, e.g. for foo.ml:
107
<< ocamlc -pp "camlp4o -I {path to type_conv} \
108
-I {path to sexplib} pa_type_conv.cmo pa_sexp_conv.cmo" \
109
-I {path to sexplib} foo.ml
112
If you are using OCamlMakefile, just put the following line at the top of
113
the file, assuming you have installed both 'type_conv' and 'sexplib' with
114
ocamlfind. The comment must start at the beginning of the line, and you must
115
not break lines (here broken for readability only):
116
<< (*pp camlp4o -I `ocamlfind query type_conv` \
117
-I `ocamlfind query sexplib` \
118
pa_type_conv.cmo pa_sexp_conv.cmo *)
121
Don't forget to place the macro 'TYPE_CONV_PATH', which takes a string
122
argument, at the top of the file to be preprocessed (see example
123
'conv_test.ml'). It is supposed to set the module path to the given module
124
file. This is necessary, because modules may get packed at a later stage, and
125
error messages generated by Sexplib will refer to this location to help
126
pinpointing the associated type.
128
In the linking stage you will only have to link with 'sexplib'. E.g. when
129
using OCamlMakefile just add it to the 'PACKS'-variable.
132
4 Syntax Specification of S-expressions
133
*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
138
4.1 Lexical conventions of S-expression
139
========================================
141
Whitespace, which consists of space, newline, carriage return, horizontal
142
tab and form feed, is ignored unless within an OCaml-string, where it is
143
treated according to OCaml-conventions. The semicolon introduces comments.
144
Comments are ignored, and range up to the next newline character. The left
145
parenthesis opens a new list, the right parenthesis closes it again. Lists can
146
be empty. The double quote denotes the beginning and end of a string following
147
the lexical conventions of OCaml (see OCaml-manual for details). All
148
characters other than double quotes, left- and right parentheses, and
149
whitespace are considered part of a contiguous string.
152
4.2 Grammar of S-expressions
153
=============================
155
S-expressions are either strings (= atoms) or lists. The lists can
156
recursively contain further S-expressions or be empty, and must be balanced,
157
i.e. parentheses must match.
163
<< this_is_an_atom_123'&^%! ; this is a comment
164
"another atom in an OCaml-string \"string in a string\" \123"
166
; empty list follows below
169
; a more complex example
172
list in a list ; comment within a list
173
(list in a list in a list)
174
42 is the answer to all questions
181
4.4 Conversion of basic OCaml-values
182
=====================================
184
Basic OCaml-values like the unit-value, integers (in all representations),
185
floats, strings, and booleans are represented in S-exp syntax in the same way
186
as in OCaml. Strings may also appear without quotes if this does not clash
187
with the lexical conventions for S-expressions.
190
4.5 Conversion of OCaml-tuples
191
===============================
193
OCaml-tuples are simple lists of values in the same order as in the tuple.
195
<< (3.14, "foo", "bar bla", 27) <===> (3.14 foo "bar bla" 27)
200
4.6 Conversion of OCaml-records
201
================================
203
OCaml-records are represented as lists of pairs in S-expression syntax. Each
204
pair consists of the name of the record field (first element), and its value
205
(second element). E.g.:
219
Type specifications of records allow the use of a special type 'sexp_option'
220
which indicates that a record field should be optional. E.g.:
228
The type 'sexp_option' is equivalent to ordinary options, but is treated
229
specially by the code generator. The above would lead to the following
230
equivalences of values and S-expressions:
257
Note how 'sexp_option' allows you to leave away record fields that should
258
default to 'None'. It is also unnecessary (and actually wrong) now to write
259
down such a value as an option, i.e. the 'some'-tag must be dropped if the
260
field should be defined.
262
The types 'sexp_list', 'sexp_array', and 'sexp_bool' can be used in ways
263
similar to the type 'sexp_option'. They assume the empty list, empty array,
264
and false value respectively as default values.
267
4.7 Conversion of sum types
268
============================
270
Constant constructors in sum types are represented as strings. Constructors
271
with arguments are represented as lists, the first element being the
272
constructor name, the rest being its arguments. Constructors may also be
273
started in lowercase in S-expressions, but will always be converted to
274
uppercase when converting from OCaml-values.
277
<< type t = A | B of int * float * t with sexp
279
B (42, 3.14, B (-1, 2.72, A)) <===> (B 42 3.14 (B -1 2.72 A))
282
The above example also demonstrates recursion in datastructures.
285
4.8 Conversion of variant types
286
================================
288
The conversion of polymorphic variants is almost the same as with sum types.
289
The notable difference is that variant constructors must always start with an
290
either lower- or uppercase character, matching the way it was specified in the
291
type definition. This is because OCaml also distinguishes between upper- and
292
lowercase variant constructors. Note that type specifications containing
293
unions of variant types are also supported by the S-expression converter.
296
4.9 Conversion of OCaml-lists and arrays
297
=========================================
299
OCaml-lists and arrays are straightforwardly represented as S-expression
303
4.10 Conversion of option types
304
================================
306
The option type is converted like ordinary polymorphic sum types, i.e.:
308
Some value <===> (some value)
311
There is a deprecated version of the syntax in which values of option type
312
are represented as lists in S-expressions:
314
Some value <===> (value)
317
Reading of the old-style S-expression syntax for option values is only
318
supported if the reference 'Conv.read_old_option_format' is set to 'true'
319
(currently the default, which may change soon). A conversion exception is
320
raised otherwise. The old format will be written only if
321
'Conv.write_old_option_format' is true (also currently the default). Reading
322
of the new format is always supported.
325
4.11 Conversion of polymorphic values
326
======================================
328
There is nothing special about polymorphic values as long as there are
329
conversion functions for the type parameters. E.g.:
330
<< type 'a t = A | B of 'a with sexp
331
type foo = int t with sexp
334
In the above case the conversion functions will behave as if 'foo' had been
335
defined as a monomorphic version of 't' with ''a' replaced by 'int' on the
338
If a datastructure is indeed polymorphic, and you want to convert it, you
339
will have to supply the conversion functions for the type parameters at
340
runtime. E.g. in the above example, if you wanted to convert a value of type
341
''a t', you would have to write something like this:
342
<< sexp_of_t sexp_of_a v
345
where 'sexp_of_a', which may also be named differently in this particular
346
case, is a function that converts values of type ''a' to an S-expression.
347
Types with more than one parameter require passing conversion functions for
348
those parameters in the order of their appearance on the left hand side of the
352
4.12 Conversion of abstract datatypes
353
======================================
355
Of course, if you want to convert an abstract datatype to an S-expression,
356
you will have to roll your own conversion function, which should produce
357
values of type 'Sexp.t' directly. If, however, you want to make use of your
358
abstract type within definitions of other types, make sure that you call your
359
conversion function appropriately: it should be in the same scope as the
360
typename, and must be named 'sexp_of_{typename}'.
363
4.13 Conversion of hashtables
364
==============================
366
Hashtables, which are abstract values in OCaml, are represented as
367
association lists, i.e. lists of key-value pairs, e.g.:
368
<< ((foo 42) (bar 3))
371
Reading in the above S-expression as hashtable mapping strings to integers
372
('(string, int) Hashtbl.t') will map '"foo"' to 42 and '"bar"' to 3.
374
Note that the order of elements in the list may matter, because duplicates
375
are kept: bindings will be inserted into the hashtable in order of appearence.
376
Therefore, the last binding of a key will be the "visible" one, the others are
377
"hidden". See the OCaml-documentation on hashtables for details.
379
Note, too, that polymorphic equality may not hold between conversions. You
380
will have to use a function implementing logical equality for that purpose.
383
4.14 Conversion of opaque values
384
=================================
386
Opaque values are ones for which we do not want to perform conversions. This
387
may be, because we do not have S-expression converters for them, or because we
388
do not want to apply them in a particular type context, e.g. if the resulting
389
S-expression should be printed out but without superfluous information. To
390
prevent the preprocessor from generating calls to converters, simply apply the
391
qualifier 'sexp_opaque' as if it were a type constructor, e.g.:
392
<< type foo = int * stuff sexp_opaque with sexp
395
Thus, there is no need to specify converters for type 'stuff', and if there
396
are any, they will not be used in this particular context. Needless to say, it
397
is not possible to convert such an S-expression back to the original value.
398
Here is an example conversion:
399
<< (42, some_stuff) ===> (42, <opaque>)
404
4.15 Conversion of exceptions
405
==============================
407
S-expression converters for exceptions can be automatically registered using
408
the 'with sexp' macro, e.g.:
410
exception Foo of int with sexp
414
Such exceptions will be translated in a similar way as sum types, but their
415
constructor will be prefixed with the fully qualified module path (here:
416
'M.Foo') so as to be able to discriminate between them without problems.
417
The user can then easily convert an exception matching the above one to an
418
S-expression using 'Sexplib.Conv.sexp_of_exn'. User-defined conversion
419
functions can be registered, too, by calling 'Sexplib.Conv.add_exn_converter'.
420
This should make it very convenient for users to catch arbitrary exceptions
421
escaping their program and pretty-printing them, including all arguments, as
422
S-expressions. The library already contains mappings for all known exceptions
423
that can escape functions in the OCaml standard library.
426
5 I/O and type conversions
427
*=*=*=*=*=*=*=*=*=*=*=*=*=*
430
There are multiple ways of performing I/O with S-expressions. If exact error
431
locations are required when type conversions fail, S-expressions need to be
432
parsed with location annotations. In most cases users may want to use
433
functions like e.g. 'load_sexp_conv' or 'load_sexp_conv_exn', which load
434
S-expressions from files and convert them. Only when conversions fail, the
435
file will be reparsed with annotations, which is slower, and type errors will
436
be reported accurately with file, line number, column, and file position.
439
6 Contact information
440
*=*=*=*=*=*=*=*=*=*=*=
443
In the case of bugs, feature requests and similar, you can contact us here:
444
opensource@janestreet.com
445
Up-to-date information concerning this library should be available here:
446
http://www.janestreet.com/ocaml
449
-----------------------------------------------------------------------------
451
This document was translated from LaTeX by HeVeA (2).
452
--------------------------------------
455
(1) http://www.janestreet.com
457
(2) http://hevea.inria.fr/index.html