auto_syntax.py

import binascii
import re


legacy_syntax_names = {
    ('\t', 'simple'): ('TSV (Rainbow)', 'tsv'),
    (',', 'quoted'): ('CSV (Rainbow)', 'csv'),
}


filename_policy_map = {'simple': 'Simple', 'quoted': 'Standard', 'quoted_rfc': 'quoted_rfc'}


def encode_delim(delim):
    return binascii.hexlify(delim.encode('utf-8')).decode('ascii')


def decode_delim(delim):
    return binascii.unhexlify(delim.encode('ascii')).decode('utf-8')


def get_syntax_file_basename(delim, policy):
    for k, (v, _ext) in legacy_syntax_names.items():
        if (delim, policy) == k:
            return v + '.sublime-syntax'
    return 'Rainbow_CSV_hex_{}_{}.sublime-syntax'.format(encode_delim(delim), filename_policy_map[policy])


def get_syntax_file_ext(delim, policy):
    for k, (_v, ext) in legacy_syntax_names.items():
        if k == (delim, policy):
            return ext
    return None


simple_header_template = '''%YAML 1.2
---
name: '{}'
file_extensions: [{}]
scope: text.csv.{}


contexts:
    main:
        - match: '^'
          push: rainbow1
'''


standard_header_template = '''%YAML 1.2
---
name: '{}'
file_extensions: [{}]
scope: text.csv.{}


contexts:
    main:
        - match: '^'
          push: rainbow1

    quoted_field:
        - match: '""'
          scope: meta.rainbow.double-quote-escaped
        - match: '"'
          pop: true
'''

non_rfc_endline_rule = '''        - match: '$'\n          pop: true\n'''


rainbow_scope_names = [
    'rainbow1',
    'keyword.rainbow2',
    'entity.name.rainbow3',
    'comment.rainbow4',
    'string.rainbow5',
    'entity.name.tag.rainbow6',
    'storage.type.rainbow7',
    'support.rainbow8',
    'constant.language.rainbow9',
    'variable.language.rainbow10'
]


def oniguruma_regular_escape_single_char(delim_char):
    single_escape_chars = r'\/|.$^*()[]+?'
    if single_escape_chars.find(delim_char) != -1:
        return r'\{}'.format(delim_char)
    if delim_char == '\t':
        return r'\t'
    return delim_char


def oniguruma_regular_escape(delim):
    return ''.join([oniguruma_regular_escape_single_char(d) for d in delim])


def get_syntax_name(delim, policy):
    for k, (v, _ext) in legacy_syntax_names.items():
        if (delim, policy) == k:
            return v
    ui_delim = delim.replace('\t', 'tab')

    hr_policy_map = {'simple': 'Simple', 'quoted': 'Standard', 'quoted_rfc': 'RFC'}
    return 'Rainbow CSV {} {}'.format(ui_delim, hr_policy_map[policy])


def yaml_escape(data):
    return data.replace("'", "''")


def get_context_name(context_id):
    return "rainbow{}".format(context_id + 1)


def make_simple_context(delim, context_id, num_contexts, indent='    '):
    result_lines = []
    next_context_id = (context_id + 1) % num_contexts
    context_header = "{}:".format(get_context_name(context_id))
    # We use `meta_content_scope` instead of `meta_scope` to prevent wrong separator color bug, see https://github.com/mechatroner/sublime_rainbow_csv/issues/31
    result_lines.append("- meta_content_scope: {}".format(rainbow_scope_names[context_id]))
    result_lines.append("- match: '{}'".format(yaml_escape(oniguruma_regular_escape(delim))))
    result_lines.append("  set: {}".format(get_context_name(next_context_id)))
    result_lines.append("- match: '$'")
    result_lines.append("  pop: true")
    result_lines = [indent + v for v in result_lines]
    result_lines = [context_header] + result_lines
    result_lines = [indent + v for v in result_lines]
    return '\n'.join(result_lines) + '\n'


def make_standard_context(delim, context_id, num_contexts, indent='    '):
    result_lines = []
    next_context_id = (context_id + 1) % num_contexts
    context_header = "{}:".format(get_context_name(context_id))
    # We use `meta_content_scope` instead of `meta_scope` to prevent wrong separator color bug, see https://github.com/mechatroner/sublime_rainbow_csv/issues/31
    result_lines.append("- meta_content_scope: {}".format(rainbow_scope_names[context_id]))
    result_lines.append("- match: '{}'".format(yaml_escape(oniguruma_regular_escape(delim))))
    result_lines.append("  set: {}".format(get_context_name(next_context_id)))
    result_lines.append("- match: '$'")
    result_lines.append("  pop: true")
    result_lines.append("- match: '\"'")
    result_lines.append("  push: quoted_field")
    result_lines = [indent + v for v in result_lines]
    result_lines = [context_header] + result_lines
    result_lines = [indent + v for v in result_lines]
    return '\n'.join(result_lines) + '\n'


def make_sublime_syntax_simple(delim):
    scope = 'rbcsmn' + ''.join([str(ord(d)) for d in delim])
    name = get_syntax_name(delim, 'simple')
    ext = get_syntax_file_ext(delim, 'simple') or scope
    result = simple_header_template.format(yaml_escape(name), ext, scope)
    num_contexts = len(rainbow_scope_names)
    for context_id in range(num_contexts):
        result += '\n'
        result += make_simple_context(delim, context_id, num_contexts)
    return result


def make_sublime_syntax_standard(delim, policy):
    assert policy in ['quoted', 'quoted_rfc']
    scope = 'rbcstn' + ''.join([str(ord(d)) for d in delim])
    name = get_syntax_name(delim, policy)
    ext = get_syntax_file_ext(delim, policy) or scope
    result = standard_header_template.format(yaml_escape(name), ext, scope)
    if policy == 'quoted':
        result += non_rfc_endline_rule
    num_contexts = len(rainbow_scope_names)
    for context_id in range(num_contexts):
        result += '\n'
        result += make_standard_context(delim, context_id, num_contexts)
    return result


def make_sublime_syntax(delim, policy):
    assert policy in filename_policy_map.keys()
    if policy == 'quoted':
        return make_sublime_syntax_standard(delim, policy)
    elif policy == 'quoted_rfc':
        return make_sublime_syntax_standard(delim, policy)
    else:
        return make_sublime_syntax_simple(delim)


def get_pregenerated_delims():
    delims = [chr(i) for i in range(32, 127)]
    delims.append('\t')
    delims = [delim for delim in delims if re.match('^[a-zA-Z0-9]$', delim) is None]
    return delims