From 3d4fb02e2907b2532333fdb5eefe6335c7cd94c4 Mon Sep 17 00:00:00 2001
From: Paolo Bonzini <pbonzini@redhat.com>
Date: Mon, 17 Aug 2020 15:26:11 +0200
Subject: ninjabackend: optimize ninja_quote

Use regular expressions to quickly weed out strings that require quoting
On a QEMU build the time spent in ninja_quote goes from 1.978s to 1.281s,
with str.replace being kicked completely out of the profile.
---
 mesonbuild/backend/ninjabackend.py | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

(limited to 'mesonbuild/backend/ninjabackend.py')

diff --git a/mesonbuild/backend/ninjabackend.py b/mesonbuild/backend/ninjabackend.py
index b4ebdc3..5f7f03a 100644
--- a/mesonbuild/backend/ninjabackend.py
+++ b/mesonbuild/backend/ninjabackend.py
@@ -114,13 +114,17 @@ rsp_threshold = get_rsp_threshold()
 # from, etc.), so it must not be shell quoted.
 raw_names = {'DEPFILE_UNQUOTED', 'DESC', 'pool', 'description', 'targetdep'}
 
+NINJA_QUOTE_BUILD_PAT = re.compile(r"[$ :\n]")
+NINJA_QUOTE_VAR_PAT = re.compile(r"[$ \n]")
+
 def ninja_quote(text, is_build_line=False):
     if is_build_line:
-        qcs = ('$', ' ', ':')
+        quote_re = NINJA_QUOTE_BUILD_PAT
     else:
-        qcs = ('$', ' ')
-    for char in qcs:
-        text = text.replace(char, '$' + char)
+        quote_re = NINJA_QUOTE_VAR_PAT
+    # Fast path for when no quoting is necessary
+    if not quote_re.search(text):
+        return text
     if '\n' in text:
         errmsg = '''Ninja does not support newlines in rules. The content was:
 
@@ -128,7 +132,7 @@ def ninja_quote(text, is_build_line=False):
 
 Please report this error with a test case to the Meson bug tracker.'''.format(text)
         raise MesonException(errmsg)
-    return text
+    return quote_re.sub(r'$\g<0>', text)
 
 @unique
 class Quoting(Enum):
-- 
cgit v1.1


From b14601da7b42f2b68bb098fc04e267a390f201c2 Mon Sep 17 00:00:00 2001
From: Paolo Bonzini <pbonzini@redhat.com>
Date: Mon, 17 Aug 2020 15:33:12 +0200
Subject: ninjabackend: optimize length_estimate

Optimize the regular expression so that the variable expansion part
always ends up in group 1, and the trailer after the variable is discarded
in the same match.  Do not use re.sub to remove braces, and do not bother
building the expanded command, just adjust the estimated length on the fly.

functools.reduce is extremely slow, so I am keeping ' '.join(chunk).

On a QEMU build the time spend in the function goes from 1.072s to 0.757s.
---
 mesonbuild/backend/ninjabackend.py | 22 ++++++++++++----------
 1 file changed, 12 insertions(+), 10 deletions(-)

(limited to 'mesonbuild/backend/ninjabackend.py')

diff --git a/mesonbuild/backend/ninjabackend.py b/mesonbuild/backend/ninjabackend.py
index 5f7f03a..fd8be00 100644
--- a/mesonbuild/backend/ninjabackend.py
+++ b/mesonbuild/backend/ninjabackend.py
@@ -265,18 +265,20 @@ class NinjaRule:
 
         # expand variables in command
         command = ' '.join([self._quoter(x) for x in self.command + self.args])
-        expanded_command = ''
-        for m in re.finditer(r'(\${\w*})|(\$\w*)|([^$]*)', command):
-            chunk = m.group()
-            if chunk.startswith('$'):
-                chunk = chunk[1:]
-                chunk = re.sub(r'{(.*)}', r'\1', chunk)
-                chunk = ninja_vars.get(chunk, [])  # undefined ninja variables are empty
-                chunk = ' '.join(chunk)
-            expanded_command += chunk
+        estimate = len(command)
+        for m in re.finditer(r'(\${\w*}|\$\w*)?[^$]*', command):
+            if m.start(1) != -1:
+                estimate -= m.end(1) - m.start(1) + 1
+                chunk = m.group(1)
+                if chunk[1] == '{':
+                    chunk = chunk[2:-1]
+                else:
+                    chunk = chunk[1:]
+                chunk = ninja_vars.get(chunk, []) # undefined ninja variables are empty
+                estimate += len(' '.join(chunk))
 
         # determine command length
-        return len(expanded_command)
+        return estimate
 
 class NinjaBuildElement:
     def __init__(self, all_outputs, outfilenames, rulename, infilenames, implicit_outs=None):
-- 
cgit v1.1


From cbde13850f1334ef3019288096ccd95dfcdaec6e Mon Sep 17 00:00:00 2001
From: Paolo Bonzini <pbonzini@redhat.com>
Date: Mon, 17 Aug 2020 15:33:39 +0200
Subject: ninjabackend: avoid lambdas

The lambda in NinjaBuildElement.write is quite expensive, totalling 0.3s
just to do a couple function calls.  Since it is used just once, simply
inline it.

On a QEMU build, the total time spent in write from this series goes from
5.321s to 3.238s, though part of it can be attributed to previous patches.
---
 mesonbuild/backend/ninjabackend.py | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)

(limited to 'mesonbuild/backend/ninjabackend.py')

diff --git a/mesonbuild/backend/ninjabackend.py b/mesonbuild/backend/ninjabackend.py
index fd8be00..3a5c102 100644
--- a/mesonbuild/backend/ninjabackend.py
+++ b/mesonbuild/backend/ninjabackend.py
@@ -386,10 +386,9 @@ class NinjaBuildElement:
             newelems = []
             for i in elems:
                 if not should_quote or i == '&&': # Hackety hack hack
-                    quoter = ninja_quote
+                    newelems.append(ninja_quote(i))
                 else:
-                    quoter = lambda x: ninja_quote(qf(x))
-                newelems.append(quoter(i))
+                    newelems.append(ninja_quote(qf(i)))
             line += ' '.join(newelems)
             line += '\n'
             outfile.write(line)
-- 
cgit v1.1