rewrite-edges-cstruct.py

import sys
import re

contents = sys.stdin.read()
# Here is a sample input:
# contents = """
#         {&getEdgesDouble, 0.00000000000000000, boost::none, boost::none, 1, {
#             "1",
#             "10",
#             "100",
#             "1000",
#             "10000",
#             "100000",
#             "1000000",
#             "10000000",
#             "100000000",
#             "1000000000",
#             "10000000000",
#             "100000000000",
#             "1000000000000",
#             "10000000000000",
#             "100000000000000",
#             "1000000000000000",
#             "10000000000000000",
#             "100000000000000000",
#             "1000000000000000000",
#             "10000000000000000000",
#             "100000000000000000000",
#             "1000000000000000000000",
#             "10000000000000000000000",
#             "100000000000000000000000",
#             "1000000000000000000000000",
#             "10000000000000000000000000",
#             "100000000000000000000000000",
#             "1000000000000000000000000000",
#             "10000000000000000000000000000",
#             "100000000000000000000000000000",
#             "1000000000000000000000000000000",
#             "10000000000000000000000000000000",
#             "100000000000000000000000000000000",
#             "1000000000000000000000000000000000",
#             "10000000000000000000000000000000000",
#             "100000000000000000000000000000000000",
#             "1000000000000000000000000000000000000",
#             "10000000000000000000000000000000000000",
#             "100000000000000000000000000000000000000",
#             "1000000000000000000000000000000000000000",
#             "10000000000000000000000000000000000000000",
#             "100000000000000000000000000000000000000000",
#             "1000000000000000000000000000000000000000000",
#             "10000000000000000000000000000000000000000000",
#             "100000000000000000000000000000000000000000000",
#             "1000000000000000000000000000000000000000000000",
#             "10000000000000000000000000000000000000000000000",
#             "100000000000000000000000000000000000000000000000",
#             "1000000000000000000000000000000000000000000000000",
#             "10000000000000000000000000000000000000000000000000",
#             "100000000000000000000000000000000000000000000000000",
#             "1000000000000000000000000000000000000000000000000000",
#             "10000000000000000000000000000000000000000000000000000",
#             "100000000000000000000000000000000000000000000000000000",
#             "1000000000000000000000000000000000000000000000000000000",
#             "10000000000000000000000000000000000000000000000000000000",
#             "100000000000000000000000000000000000000000000000000000000",
#             "1000000000000000000000000000000000000000000000000000000000",
#             "10000000000000000000000000000000000000000000000000000000000",
#             "100000000000000000000000000000000000000000000000000000000000",
#             "1000000000000000000000000000000000000000000000000000000000000",
#             "10000000000000000000000000000000000000000000000000000000000000",
#             "100000000000000000000000000000000000000000000000000000000000000",
#             "1000000000000000000000000000000000000000000000000000000000000000",
#             "root",
#         }},
# """
lines = contents.splitlines ()
edges = []
import textwrap

tmpl = """
{{
    .value = {value},
    .sparsity = {sparsity},
    .expectEdges = {{
{expectEdgesString}
    }}
}},
"""

print ("""// This is a copy of test vectors from the mongodb/mongo repository.
// The order of edges is rearranged to permit exact matching.
""")



lineno = 0
import sys
for line in lines:
    lineno += 1
    match = re.match(r".*\{.+, (.+), (.+), (.+), ([-0-9]+),", line)
    if match != None:
        value = match.group(1)
        min = match.group(2)
        max = match.group(3)
        sparsity = match.group(4)

    match = re.match (r'.*"(.*)",', line)
    if match != None:
        edges.append('"' + match.group(1) + '"')
    elif re.match (r'.*\}\},', line) != None:
        # Rewrite edges. "root" should be first. Then leaf. Then remaining.
        edges.insert(0, edges[-2])
        edges.insert(0, '"root"')
        edges = edges[0:-2]

        if value is None:
            print (f"value is unexpectedly None on line {lineno}")
            sys.exit(1)

        if sparsity is None:
            print (f"sparsity is unexpectedly None on line {lineno}")
            sys.exit(1)
        
        if value == "std::numeric_limits<double>::min()":
            value = "DBL_MIN"

        if value == "std::numeric_limits<double>::max()":
            value = "DBL_MAX"
        
        got = tmpl.format(
            value=value,
            sparsity=sparsity,
            expectEdgesString= textwrap.indent (",\n".join(edges), "        ")
        )
        print(got.strip())
        edges = []
        value = None
        min = None
        max = None
        sparsity = None