Fix #323: allow identifiers on EDS native codec

CHANGELOG.md

@@ -5,6 +5,7 @@
 ### Added
 
 * `delphin.eds.EDSWarning`
+* `delphin.codecs.eds` now reads and writes identifiers ([#323])
 
 ### Fixed
 
@@ -1462,3 +1463,4 @@ information about changes, except for
 [#306]: https://github.com/delph-in/pydelphin/issues/306
 [#316]: https://github.com/delph-in/pydelphin/issues/316
 [#319]: https://github.com/delph-in/pydelphin/issues/319
+[#323]: https://github.com/delph-in/pydelphin/issues/323

delphin/codecs/eds.py

@@ -138,6 +138,7 @@ def encode(e, properties=True, lnk=True, show_status=False, indent=False):
 
 _EDSLexer = Lexer(
     tokens=[
+        (r'\#([^\s\{]+)\s*(?=\{|$)', 'IDENTIFIER'),
         (r'\{', 'LBRACE:{'),
         (r'\}', 'RBRACE:}'),
         (r'\((?:cyclic *)?(?:fragmented)?\)', 'GRAPHSTATUS'),
@@ -153,6 +154,7 @@ def encode(e, properties=True, lnk=True, show_status=False, indent=False):
     ],
     error_class=EDSSyntaxError)
 
+IDENTIFIER  = _EDSLexer.tokentypes.IDENTIFIER
 LBRACE      = _EDSLexer.tokentypes.LBRACE
 RBRACE      = _EDSLexer.tokentypes.RBRACE
 GRAPHSTATUS = _EDSLexer.tokentypes.GRAPHSTATUS
@@ -176,6 +178,7 @@ def _decode(lineiter):
 
 
 def _decode_eds(lexer):
+    identifier = lexer.accept_type(IDENTIFIER)
     lexer.expect_type(LBRACE)
 
     # after the LBRACE, the following patterns determine the top:
@@ -204,7 +207,7 @@ def _decode_eds(lexer):
         start, _ = lexer.expect_type(SYMBOL, COLON)
         nodes.append(_decode_node(start, lexer))
     lexer.expect_type(RBRACE)
-    return EDS(top=top, nodes=nodes)
+    return EDS(top=top, nodes=nodes, identifier=identifier)
 
 
 def _decode_node(start, lexer):
@@ -249,9 +252,14 @@ def _decode_edges(start, lexer):
 # Encoding
 
 def _encode_eds(e, properties, lnk, show_status, indent):
+    start = '{'
+    if e.identifier:
+        start = f'#{e.identifier}' + ('\n' if indent else ' ') + '{'
+    end = '\n}' if indent else '}'
+
     # do something predictable for empty EDS
     if len(e.nodes) == 0:
-        return '{\n}' if indent else '{}'
+        return start + end
 
     delim = '\n' if indent else ' '
     connected = ' ' if indent else ''
@@ -279,7 +287,7 @@ def _encode_eds(e, properties, lnk, show_status, indent):
         membership = connected if node.id in nidgrp else disconnected
         parts.append(membership + _encode_node(node, properties, lnk))
 
-    return '{' + delim.join(parts) + ('\n}' if indent else '}')
+    return start + delim.join(parts) + end
 
 
 def _encode_node(node, properties, lnk):

tests/codecs/edsnative_test.py

@@ -68,7 +68,23 @@ def test_decode_no_top():
     assert len(e.nodes) == 3
 
 
+def test_decode_identifier():
+    e = edsnative.decode(
+        '#123\n'
+        '{e2:\n'
+        ' e2:_rain_v_1<3:9>{e SF prop, TENSE pres}[]\n'
+        '}'
+    )
+    assert e.identifier == '123'
+    e = edsnative.decode(
+        '#123 {e2: e2:_rain_v_1<3:9>{e SF prop, TENSE pres}[] }'
+    )
+    assert e.identifier == '123'
+
+
 def test_encode(dogs_bark_from_mrs):
+    assert edsnative.encode(EDS()) == '{}'
+    assert edsnative.encode(EDS(), indent=True) == '{\n}'
     d = EDS(**dogs_bark_from_mrs)
     assert edsnative.encode(d) == (
         '{e2: e2:_bark_v_1{e}[ARG1 x4] _1:udef_q[BV x4] x4:_dog_n_1{x}[]}')
@@ -91,3 +107,18 @@ def test_encode_no_top(dogs_bark_from_mrs):
         ' _1:udef_q[BV x4]\n'
         ' x4:_dog_n_1{x}[]\n'
         '}')
+
+
+def test_encode_identifier(dogs_bark_from_mrs):
+    assert edsnative.encode(EDS(identifier='123')) == '#123 {}'
+    d = EDS(**dogs_bark_from_mrs)
+    d.identifier = '123'
+    assert edsnative.encode(d) == (
+        '#123 {e2: e2:_bark_v_1{e}[ARG1 x4] _1:udef_q[BV x4] x4:_dog_n_1{x}[]}')
+    assert edsnative.encode(d, indent=True) == (
+        '#123\n'
+        '{e2:\n'
+        ' e2:_bark_v_1{e}[ARG1 x4]\n'
+        ' _1:udef_q[BV x4]\n'
+        ' x4:_dog_n_1{x}[]\n'
+        '}')