json: merge lit sequences and handle negatives
{"type": "string", "pattern": "^({\"question\": \"[^\"]+\", \"response\": \"[^\"]+\"}\\n)+$"}
This commit is contained in:
parent
a78eb4a0c3
commit
d5ef412f31
1 changed files with 20 additions and 3 deletions
|
@ -1,5 +1,6 @@
|
||||||
#!/usr/bin/env python3
|
#!/usr/bin/env python3
|
||||||
import argparse
|
import argparse
|
||||||
|
import itertools
|
||||||
import json
|
import json
|
||||||
import re
|
import re
|
||||||
import sys
|
import sys
|
||||||
|
@ -76,9 +77,25 @@ class SchemaConverter:
|
||||||
assert pattern.startswith('^') and pattern.endswith('$'), 'Pattern must start with "^" and end with "$"'
|
assert pattern.startswith('^') and pattern.endswith('$'), 'Pattern must start with "^" and end with "$"'
|
||||||
pattern = pattern[1:-1]
|
pattern = pattern[1:-1]
|
||||||
try:
|
try:
|
||||||
|
def visit_seq(seq):
|
||||||
|
out = []
|
||||||
|
for t, g in itertools.groupby(seq, lambda x: x[0]):
|
||||||
|
g = list(g)
|
||||||
|
if t == re._parser.LITERAL and len(g) > 1:
|
||||||
|
out.append(self._format_literal(''.join(chr(x[1]) for x in g)))
|
||||||
|
else:
|
||||||
|
out.extend(visit(x) for x in g)
|
||||||
|
if len(out) == 1:
|
||||||
|
return out[0]
|
||||||
|
return '(' + ' '.join(out) + ')'
|
||||||
|
|
||||||
def visit(pattern):
|
def visit(pattern):
|
||||||
if pattern[0] == re._parser.LITERAL:
|
if pattern[0] == re._parser.LITERAL:
|
||||||
return json.dumps(chr(pattern[1]))
|
return json.dumps(chr(pattern[1]))
|
||||||
|
elif pattern[0] == re._parser.NOT_LITERAL:
|
||||||
|
ch = chr(pattern[1])
|
||||||
|
esc_ch = '\\' + ch if ch in ('-', ']', '\\') else ch
|
||||||
|
return f'[^{esc_ch}]'
|
||||||
elif pattern[0] == re._parser.ANY:
|
elif pattern[0] == re._parser.ANY:
|
||||||
raise ValueError('Unsupported pattern: "."')
|
raise ValueError('Unsupported pattern: "."')
|
||||||
elif pattern[0] == re._parser.IN:
|
elif pattern[0] == re._parser.IN:
|
||||||
|
@ -112,9 +129,9 @@ class SchemaConverter:
|
||||||
else:
|
else:
|
||||||
raise ValueError(f'Unrecognized pattern: {pattern} ({type(pattern)}; min: {min_times}, max: {max_times})')
|
raise ValueError(f'Unrecognized pattern: {pattern} ({type(pattern)}; min: {min_times}, max: {max_times})')
|
||||||
elif isinstance(pattern, re._parser.SubPattern):
|
elif isinstance(pattern, re._parser.SubPattern):
|
||||||
return ' '.join(visit(p) for p in pattern.data)
|
return visit_seq(pattern.data)
|
||||||
elif isinstance(pattern, list):# and (len(pattern) == 0 or isinstance(pattern[0], (tuple, list))):
|
elif isinstance(pattern, list):
|
||||||
return ' '.join(visit(p) for p in pattern)
|
return visit_seq(pattern)
|
||||||
else:
|
else:
|
||||||
raise ValueError(f'Unrecognized pattern: {pattern} ({type(pattern)})')
|
raise ValueError(f'Unrecognized pattern: {pattern} ({type(pattern)})')
|
||||||
return visit(re._parser.parse(pattern))
|
return visit(re._parser.parse(pattern))
|
||||||
|
|
Loading…
Add table
Add a link
Reference in a new issue