Patchwork D11682: dirstate-v2: adds two flag to track the presence of some unrecorded files

login
register
mail settings
Submitter phabricator
Date Oct. 18, 2021, 6:19 a.m.
Message ID <differential-rev-PHID-DREV-vsj6veyf3f2ptkwsbxpw-req@mercurial-scm.org>
Download mbox | patch
Permalink /patch/50005/
State Superseded
Headers show

Comments

phabricator - Oct. 18, 2021, 6:19 a.m.
marmoute created this revision.
Herald added a reviewer: hg-reviewers.
Herald added a subscriber: mercurial-patches.

REVISION SUMMARY
  Right now, we don't record ignored or unknown files in the dirstate. However the
  structure would allow it. So we introduce two flags that can be used to clarify
  why unknown/ignored file might be missing. This will allow for more information
  to be stored in the future if this end up being relevant.

REPOSITORY
  rHG Mercurial

BRANCH
  default

REVISION DETAIL
  https://phab.mercurial-scm.org/D11682

AFFECTED FILES
  mercurial/cext/util.h
  mercurial/dirstateutils/v2.py
  mercurial/helptext/internals/dirstate-v2.txt
  mercurial/pure/parsers.py
  rust/hg-core/src/dirstate_tree/on_disk.rs

CHANGE DETAILS




To: marmoute, #hg-reviewers
Cc: mercurial-patches, mercurial-devel

Patch

diff --git a/rust/hg-core/src/dirstate_tree/on_disk.rs b/rust/hg-core/src/dirstate_tree/on_disk.rs
--- a/rust/hg-core/src/dirstate_tree/on_disk.rs
+++ b/rust/hg-core/src/dirstate_tree/on_disk.rs
@@ -111,6 +111,8 @@ 
         const MODE_EXEC_PERM = 1 << 6;
         const MODE_IS_SYMLINK = 1 << 7;
         const EXPECTED_STATE_IS_MODIFIED = 1 << 8;
+        const UNRECORDED_UNKNOWN = 1 << 9;
+        const UNRECORDED_IGNORED = 1 << 10;
     }
 }
 
@@ -322,7 +324,11 @@ 
     pub(super) fn cached_directory_mtime(
         &self,
     ) -> Result<Option<TruncatedTimestamp>, DirstateV2ParseError> {
-        if self.flags().contains(Flags::HAS_DIRECTORY_MTIME) {
+        // For now we do not have code to handle UNRECORDED_UNKNOWN, so we
+        // ignore the mtime if the flag is set.
+        if self.flags().contains(Flags::HAS_DIRECTORY_MTIME)
+            && !self.flags().contains(Flags::UNRECORDED_UNKNOWN)
+        {
             if self.flags().contains(Flags::HAS_FILE_MTIME) {
                 Err(DirstateV2ParseError)
             } else {
@@ -589,7 +595,14 @@ 
                             Node::from_dirstate_entry(entry)
                         }
                         dirstate_map::NodeData::CachedDirectory { mtime } => (
-                            Flags::HAS_DIRECTORY_MTIME,
+                            // we currently never record ignored files so we
+                            // blindly set UNRECORDED_IGNORED,. We don't need
+                            // to set
+                            // UNRECORDED_UNKNOWN since we currently never set
+                            // HAS_DIRECTORY_MTIME
+                            // if unknow files were present.
+                            Flags::HAS_DIRECTORY_MTIME
+                                | Flags::UNRECORDED_IGNORED,
                             0.into(),
                             (*mtime).into(),
                         ),
diff --git a/mercurial/pure/parsers.py b/mercurial/pure/parsers.py
--- a/mercurial/pure/parsers.py
+++ b/mercurial/pure/parsers.py
@@ -54,6 +54,8 @@ 
 DIRSTATE_V2_MODE_EXEC_PERM = 1 << 6
 DIRSTATE_V2_MODE_IS_SYMLINK = 1 << 7
 DIRSTATE_V2_EXPECTED_STATE_IS_MODIFIED = 1 << 8
+DIRSTATE_V2_UNRECORDED_UNKNOWN = 1 << 9
+DIRSTATE_V2_UNRECORDED_IGNORED = 1 << 10
 
 
 @attr.s(slots=True, init=False)
@@ -340,6 +342,9 @@ 
                 flags |= DIRSTATE_V2_MODE_IS_SYMLINK
         if self._mtime is not None:
             flags |= DIRSTATE_V2_HAS_FILE_MTIME
+        # Note: we do not need to do anything regarding
+        # DIRSTATE_V2_UNRECORDED_UNKNOWN and DIRSTATE_V2_UNRECORDED_IGNORED
+        # since we never set _DIRSTATE_V2_HAS_DIRCTORY_MTIME
         return (flags, self._size or 0, self._mtime or 0)
 
     def v1_state(self):
diff --git a/mercurial/helptext/internals/dirstate-v2.txt b/mercurial/helptext/internals/dirstate-v2.txt
--- a/mercurial/helptext/internals/dirstate-v2.txt
+++ b/mercurial/helptext/internals/dirstate-v2.txt
@@ -384,6 +384,8 @@ 
     MODE_EXEC_PERM = 1 << 6
     MODE_IS_SYMLINK = 1 << 7
     EXPECTED_STATE_IS_MODIFIED = 1 << 8
+    UNRECORDED_UNKNOWN = 1 << 9
+    UNRECORDED_UNKNOWN = 1 << 10
 
   The meaning of each bit is described below.
 
@@ -530,3 +532,15 @@ 
     does not need to do the same again.
     It is valid to never set this bit,
     and consider expected metadata ambiguous if it is set.
+
+`UNRECORDED_UNKNOWN`
+    The node might have unknown children on disk that are not recorded as
+    children-nodes of this node. If this flag is set and the current operation
+    need to list the unknown files, the optimization from `HAS_DIRECTORY_MTIME`
+    cannot be used.
+
+`UNRECORDED_IGNORED`
+    The node might have ignored children on disk that are not recorded as
+    children-nodes of this node. If this flag is set and the current operation
+    need to list the ignored files, the optimization from `HAS_DIRECTORY_MTIME`
+    cannot be used.
diff --git a/mercurial/dirstateutils/v2.py b/mercurial/dirstateutils/v2.py
--- a/mercurial/dirstateutils/v2.py
+++ b/mercurial/dirstateutils/v2.py
@@ -81,6 +81,9 @@ 
     """parse <len> nodes from <data> starting at offset <start>
 
     This is used by parse_dirstate to recursively fill `map` and `copy_map`.
+
+    All directory specific information is ignored and do not need any
+    processing (HAS_DIRECTORY_MTIME, UNRECORDED_UNKNOWN, UNRECORDED_IGNORED)
     """
     for i in range(len):
         node_start = start + NODE_SIZE * i
diff --git a/mercurial/cext/util.h b/mercurial/cext/util.h
--- a/mercurial/cext/util.h
+++ b/mercurial/cext/util.h
@@ -40,6 +40,8 @@ 
 static const int dirstate_flag_mode_exec_perm = 1 << 6;
 static const int dirstate_flag_mode_is_symlink = 1 << 7;
 static const int dirstate_flag_expected_state_is_modified = 1 << 8;
+static const int dirstate_flag_unrecorded_unknown = 1 << 9;
+static const int dirstate_flag_unrecorded_ignored = 1 << 10;
 
 extern PyTypeObject dirstateItemType;
 #define dirstate_tuple_check(op) (Py_TYPE(op) == &dirstateItemType)