Use sql batch to optimize

This commit is contained in:
WQY\qiong
2023-03-16 08:30:04 +08:00
parent 1d2ac09c92
commit 04a58bb864
29 changed files with 145 additions and 105 deletions

View File

@@ -13,7 +13,7 @@ from .s8_tags import inp_in_tag_new
from .s9_demands import inp_in_demand_new
from .s10_status import inp_in_status_new
from .s11_patterns import inp_in_pattern_new
from .s12_curves import inp_in_curve_new
from .s12_curves import CURVE_TYPE_PUMP, inp_in_curve_new
from .s13_controls import inp_in_control_new
from .s14_rules import inp_in_rule_new
from .s15_energy import inp_in_energy_new
@@ -103,10 +103,12 @@ _level_4 = [
VERTICES,
]
def _get_offset(inp: str) -> dict[str, list[int]]:
def _get_offset(inp: str) -> tuple[dict[str, list[int]], bool]:
offset: dict[str, list[int]] = {}
current = ''
demand_outside = False
with open(inp) as f:
while True:
line = f.readline()
@@ -120,20 +122,43 @@ def _get_offset(inp: str) -> dict[str, list[int]]:
if s not in offset:
offset[s] = []
offset[s].append(f.tell())
current = s
break
elif line != '' and line.startswith(';') == False:
if current == DEMANDS:
demand_outside = True
return offset
return (offset, demand_outside)
def print_time(desc: str) -> None:
time = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')
def print_time(desc: str) -> datetime.datetime:
now = datetime.datetime.now()
time = now.strftime('%Y-%m-%d %H:%M:%S')
print(f"{time}: {desc}")
return now
class SQLBatch:
def __init__(self, project: str, count: int = 100) -> None:
self.batch: list[str] = []
self.project = project
self.count = count
def add(self, sql: str) -> None:
self.batch.append(sql)
if len(self.batch) == self.count:
self.flush()
def flush(self) -> None:
write(self.project, ''.join(self.batch))
self.batch.clear()
def parse_file(project: str, inp: str) -> None:
print_time(f"start reading {inp}...")
start = print_time(f'Start reading "{inp}"...')
offset = _get_offset(inp)
print_time("First scan...")
offset, demand_outside = _get_offset(inp)
levels = _level_1 + _level_2 + _level_3 + _level_4
@@ -143,15 +168,21 @@ def parse_file(project: str, inp: str) -> None:
if t[0] == _S:
sections[s] = []
pattern_desc_line = None
current_pattern = None
current_curve = None
curve_type_desc_line = None
demand_junction = None
sql_batch = SQLBatch(project)
print_time("Second scan...")
with open(inp) as f:
for s in levels:
if s not in offset:
continue
if s == DEMANDS and demand_outside == False:
continue
print_time(f"[{s}]")
is_s = _handler[s][0] == _S
@@ -177,39 +208,40 @@ def parse_file(project: str, inp: str) -> None:
if line.startswith(';'):
line = line.removeprefix(';')
if s == PATTERNS: # ;desc
pattern_desc_line = line
pass
elif s == CURVES: # ;type: desc
curve_type_desc_line = line
continue
if s == PATTERNS:
if pattern_desc_line != None:
tokens = line.split()
write(project, f"insert into _pattern (id) values ('{tokens[0]}');")
pattern_desc_line = None
elif s == CURVES:
if curve_type_desc_line != None:
type_and_desc = curve_type_desc_line.split(':')
tokens = line.split()
write(project, f"insert into _curve (id, type) values ('{tokens[0]}', '{type_and_desc[0].strip()}');")
curve_type_desc_line = None
elif s == DEMANDS:
tokens = line.split()
junction = str(tokens[0])
if demand_junction != junction:
if try_read(project, f"select * from demands where junction = '{junction}'") != None:
write(project, f"delete from demands where junction = '{junction}';")
demand_junction = junction
handler(project, line)
if current_pattern != tokens[0]:
write(project, f"insert into _pattern (id) values ('{tokens[0]}');")
current_pattern = tokens[0]
elif s == CURVES:
tokens = line.split()
if current_curve != tokens[0]:
type = CURVE_TYPE_PUMP
if curve_type_desc_line != None:
type = curve_type_desc_line.split(':')[0].strip()
write(project, f"insert into _curve (id, type) values ('{tokens[0]}', '{type}');")
current_curve = tokens[0]
curve_type_desc_line = None
if s == JUNCTIONS:
sql_batch.add(handler(line, demand_outside))
else:
sql_batch.add(handler(line))
f.seek(0)
if is_s:
handler(project, sections[s])
sql_batch.add(handler(sections[s]))
print_time(f"end reading {inp}")
sql_batch.flush()
end = print_time(f'End reading "{inp}"')
print(f"Total (in second): {(end-start).seconds}(s)")
def parse_inp(project: str, inp: str) -> None: