Class: Yajl::Parser

Inherits:
Object show all
Defined in:
ext/yajl/yajl_ext.c,
lib/yajl.rb,
ext/yajl/yajl_ext.c

Overview

This class contains methods for parsing JSON directly from an IO object. The only basic requirment currently is that the IO object respond to #read(len) and #eof? The IO is parsed until a complete JSON object has been read and a ruby object will be returned.

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(*args) ⇒ Object

call-seq: new([:symbolize_keys => true, [:allow_comments => false[, :check_utf8 => false]]])

:symbolize_keys will turn hash keys into Ruby symbols, defaults to false.

:allow_comments will turn on/off the check for comments inside the JSON stream, defaults to true.

:check_utf8 will validate UTF8 characters found in the JSON stream, defaults to true.


470
471
472
# File 'ext/yajl/yajl_ext.c', line 470

static VALUE rb_yajl_parser_init(int argc, VALUE * argv, VALUE self) {
    return self;
}

Class Method Details

.new(*args) ⇒ Object

call-seq: new([:symbolize_keys => true, [:allow_comments => false[, :check_utf8 => false]]])

:symbolize_keys will turn hash keys into Ruby symbols, defaults to false.

:allow_comments will turn on/off the check for comments inside the JSON stream, defaults to true.

:check_utf8 will validate UTF8 characters found in the JSON stream, defaults to true.


425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
# File 'ext/yajl/yajl_ext.c', line 425

static VALUE rb_yajl_parser_new(int argc, VALUE * argv, VALUE klass) {
    yajl_parser_wrapper * wrapper;
    yajl_parser_config cfg;
    VALUE opts, obj;
    int allowComments = 1, checkUTF8 = 1, symbolizeKeys = 0;

    /* Scan off config vars */
    if (rb_scan_args(argc, argv, "01", &opts) == 1) {
        Check_Type(opts, T_HASH);

        if (rb_hash_aref(opts, sym_allow_comments) == Qfalse) {
            allowComments = 0;
        }
        if (rb_hash_aref(opts, sym_check_utf8) == Qfalse) {
            checkUTF8 = 0;
        }
        if (rb_hash_aref(opts, sym_symbolize_keys) == Qtrue || rb_hash_aref(opts, sym_symbolize_names) == Qtrue) {
            symbolizeKeys = 1;
        }
    }
    cfg = (yajl_parser_config){allowComments, checkUTF8};

    obj = Data_Make_Struct(klass, yajl_parser_wrapper, yajl_parser_wrapper_mark, yajl_parser_wrapper_free, wrapper);
    wrapper->parser = yajl_alloc(&callbacks, &cfg, &rb_alloc_funcs, (void *)obj);
    wrapper->nestedArrayLevel = 0;
    wrapper->nestedHashLevel = 0;
    wrapper->objectsFound = 0;
    wrapper->symbolizeKeys = symbolizeKeys;
    wrapper->builderStack = rb_ary_new();
    wrapper->parse_complete_callback = Qnil;
    rb_obj_call_init(obj, 0, 0);
    return obj;
}

.parse(str_or_io, options = {}, read_bufsize = nil, &block) ⇒ Object

A helper method for parse-and-forget use-cases

io is the stream to parse JSON from

The options hash allows you to set two parsing options - :allow_comments and :check_utf8

:allow_comments accepts a boolean will enable/disable checks for in-line comments in the JSON stream

:check_utf8 accepts a boolean will enable/disable UTF8 validation for the JSON stream


43
44
45
# File 'lib/yajl.rb', line 43

def self.parse(str_or_io, options={}, read_bufsize=nil, &block)
  new(options).parse(str_or_io, read_bufsize, &block)
end

Instance Method Details

#<<(chunk) ⇒ Object

call-seq: parse_chunk(string_chunk)

string_chunk can be a partial or full JSON string to push on the parser.

This method will throw an exception if the on_parse_complete callback hasn't been assigned yet. The on_parse_complete callback assignment is required so the user can handle objects that have been parsed off the stream as they're found.


552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
# File 'ext/yajl/yajl_ext.c', line 552

static VALUE rb_yajl_parser_parse_chunk(VALUE self, VALUE chunk) {
    yajl_parser_wrapper * wrapper;
    unsigned int len;

    GetParser(self, wrapper);
    if (NIL_P(chunk)) {
        rb_raise(cParseError, "Can't parse a nil string.");
    }

    if (wrapper->parse_complete_callback != Qnil) {
        const char * cptr = RSTRING_PTR(chunk);
        len = (unsigned int)RSTRING_LEN(chunk);
        yajl_parse_chunk((const unsigned char*)cptr, len, wrapper->parser);
    } else {
        rb_raise(cParseError, "The on_parse_complete callback isn't setup, parsing useless.");
    }

    return Qnil;
}

#on_parse_complete=(callback) ⇒ Object

call-seq: on_parse_complete = Proc.new { |obj| … }

This callback setter allows you to pass a Proc/lambda or any other object that responds to #call.

It will pass a single parameter, the ruby object built from the last parsed JSON object


581
582
583
584
585
586
# File 'ext/yajl/yajl_ext.c', line 581

static VALUE rb_yajl_parser_set_complete_cb(VALUE self, VALUE callback) {
    yajl_parser_wrapper * wrapper;
    GetParser(self, wrapper);
    wrapper->parse_complete_callback = callback;
    return Qnil;
}

#parse(*args) ⇒ Object

call-seq:

parse(input, buffer_size=8192)
parse(input, buffer_size=8192) { |obj| ... }

input can either be a string or an IO to parse JSON from

buffer_size is the size of chunk that will be parsed off the input (if it's an IO) for each loop of the parsing process. 8192 is a good balance between the different types of streams (off disk, off a socket, etc…), but this option is here so the caller can better tune their parsing depending on the type of stream being passed. A larger read buffer will perform better for files off disk, where as a smaller size may be more efficient for reading off of a socket directly.

If a block was passed, it's called when an object has been parsed off the stream. This is especially usefull when parsing a stream of multiple JSON objects.

NOTE: you can optionally assign the on_parse_complete callback, and it will be called the same way the optional block is for this method.


495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
# File 'ext/yajl/yajl_ext.c', line 495

static VALUE rb_yajl_parser_parse(int argc, VALUE * argv, VALUE self) {
    yajl_status stat;
    yajl_parser_wrapper * wrapper;
    VALUE rbufsize, input, blk;
    unsigned int len;
    const char * cptr;

    GetParser(self, wrapper);

    /* setup our parameters */
    rb_scan_args(argc, argv, "11&", &input, &rbufsize, &blk);
    if (NIL_P(rbufsize)) {
        rbufsize = INT2FIX(READ_BUFSIZE);
    } else {
        Check_Type(rbufsize, T_FIXNUM);
    }
    if (!NIL_P(blk)) {
        rb_yajl_parser_set_complete_cb(self, blk);
    }

    if (TYPE(input) == T_STRING) {
        cptr = RSTRING_PTR(input);
        len = (unsigned int)RSTRING_LEN(input);
        yajl_parse_chunk((const unsigned char*)cptr, len, wrapper->parser);
    } else if (rb_respond_to(input, intern_io_read)) {
        VALUE parsed = rb_str_new(0, FIX2LONG(rbufsize));
        while (rb_funcall(input, intern_io_read, 2, rbufsize, parsed) != Qnil) {
            cptr = RSTRING_PTR(parsed);
            len = (unsigned int)RSTRING_LEN(parsed);
            yajl_parse_chunk((const unsigned char*)cptr, len, wrapper->parser);
        }
    } else {
        rb_raise(cParseError, "input must be a string or IO");
    }

    /* parse any remaining buffered data */
    stat = yajl_parse_complete(wrapper->parser);

    if (wrapper->parse_complete_callback != Qnil) {
        yajl_check_and_fire_callback((void *)self);
        return Qnil;
    }

    return rb_ary_pop(wrapper->builderStack);
}

#parse_chunk(chunk) ⇒ Object

call-seq: parse_chunk(string_chunk)

string_chunk can be a partial or full JSON string to push on the parser.

This method will throw an exception if the on_parse_complete callback hasn't been assigned yet. The on_parse_complete callback assignment is required so the user can handle objects that have been parsed off the stream as they're found.


552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
# File 'ext/yajl/yajl_ext.c', line 552

static VALUE rb_yajl_parser_parse_chunk(VALUE self, VALUE chunk) {
    yajl_parser_wrapper * wrapper;
    unsigned int len;

    GetParser(self, wrapper);
    if (NIL_P(chunk)) {
        rb_raise(cParseError, "Can't parse a nil string.");
    }

    if (wrapper->parse_complete_callback != Qnil) {
        const char * cptr = RSTRING_PTR(chunk);
        len = (unsigned int)RSTRING_LEN(chunk);
        yajl_parse_chunk((const unsigned char*)cptr, len, wrapper->parser);
    } else {
        rb_raise(cParseError, "The on_parse_complete callback isn't setup, parsing useless.");
    }

    return Qnil;
}