You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@lucy.apache.org by ma...@apache.org on 2013/07/17 16:12:41 UTC
[lucy-commits] [21/34] git commit: refs/heads/master - Migrate Perl host code to
IVARS.
Migrate Perl host code to IVARS.
Migrate host-specific code for Perl to use IVARS rather than access struct
members through `self`.
Project: http://git-wip-us.apache.org/repos/asf/lucy/repo
Commit: http://git-wip-us.apache.org/repos/asf/lucy/commit/811358d9
Tree: http://git-wip-us.apache.org/repos/asf/lucy/tree/811358d9
Diff: http://git-wip-us.apache.org/repos/asf/lucy/diff/811358d9
Branch: refs/heads/master
Commit: 811358d9f87133f6987eb64259f4b4bbcaafddc2
Parents: ff0ec8d
Author: Marvin Humphrey <ma...@rectangular.com>
Authored: Mon Jul 1 08:45:46 2013 -0700
Committer: Marvin Humphrey <ma...@rectangular.com>
Committed: Tue Jul 16 16:08:43 2013 -0700
----------------------------------------------------------------------
perl/xs/Lucy/Analysis/RegexTokenizer.c | 27 +++++++------
perl/xs/Lucy/Document/Doc.c | 60 ++++++++++++++++-------------
perl/xs/Lucy/Index/DocReader.c | 7 ++--
perl/xs/Lucy/Index/Inverter.c | 27 +++++++------
4 files changed, 69 insertions(+), 52 deletions(-)
----------------------------------------------------------------------
http://git-wip-us.apache.org/repos/asf/lucy/blob/811358d9/perl/xs/Lucy/Analysis/RegexTokenizer.c
----------------------------------------------------------------------
diff --git a/perl/xs/Lucy/Analysis/RegexTokenizer.c b/perl/xs/Lucy/Analysis/RegexTokenizer.c
index d175b59..86738fc 100644
--- a/perl/xs/Lucy/Analysis/RegexTokenizer.c
+++ b/perl/xs/Lucy/Analysis/RegexTokenizer.c
@@ -42,6 +42,7 @@ lucy_RegexTokenizer*
lucy_RegexTokenizer_init(lucy_RegexTokenizer *self,
const cfish_CharBuf *pattern) {
lucy_Analyzer_init((lucy_Analyzer*)self);
+ lucy_RegexTokenizerIVARS *const ivars = lucy_RegexTokenizer_IVARS(self);
#define DEFAULT_PATTERN "\\w+(?:['\\x{2019}]\\w+)*"
if (pattern) {
if (Cfish_CB_Find_Str(pattern, "\\p", 2) != -1
@@ -50,15 +51,15 @@ lucy_RegexTokenizer_init(lucy_RegexTokenizer *self,
CFISH_DECREF(self);
THROW(CFISH_ERR, "\\p and \\P constructs forbidden");
}
- self->pattern = Cfish_CB_Clone(pattern);
+ ivars->pattern = Cfish_CB_Clone(pattern);
}
else {
- self->pattern = cfish_CB_new_from_trusted_utf8(
+ ivars->pattern = cfish_CB_new_from_trusted_utf8(
DEFAULT_PATTERN, sizeof(DEFAULT_PATTERN) - 1);
}
// Acquire a compiled regex engine for matching one token.
- SV *token_re_sv = S_compile_token_re(self->pattern);
+ SV *token_re_sv = S_compile_token_re(ivars->pattern);
S_set_token_re_but_not_pattern(self, SvRV(token_re_sv));
SvREFCNT_dec(token_re_sv);
@@ -86,6 +87,7 @@ S_compile_token_re(const cfish_CharBuf *pattern) {
static void
S_set_token_re_but_not_pattern(lucy_RegexTokenizer *self, void *token_re) {
+ lucy_RegexTokenizerIVARS *const ivars = lucy_RegexTokenizer_IVARS(self);
#if (PERL_VERSION > 10)
REGEXP *rx = SvRX((SV*)token_re);
#else
@@ -102,17 +104,18 @@ S_set_token_re_but_not_pattern(lucy_RegexTokenizer *self, void *token_re) {
THROW(CFISH_ERR, "Failed to extract REGEXP from token_re '%s'",
SvPV_nolen((SV*)token_re));
}
- if (self->token_re) { ReREFCNT_dec(((REGEXP*)self->token_re)); }
- self->token_re = rx;
- (void)ReREFCNT_inc(((REGEXP*)self->token_re));
+ if (ivars->token_re) { ReREFCNT_dec(((REGEXP*)ivars->token_re)); }
+ ivars->token_re = rx;
+ (void)ReREFCNT_inc(((REGEXP*)ivars->token_re));
}
static void
S_set_pattern_from_token_re(lucy_RegexTokenizer *self, void *token_re) {
+ lucy_RegexTokenizerIVARS *const ivars = lucy_RegexTokenizer_IVARS(self);
SV *rv = newRV((SV*)token_re);
STRLEN len = 0;
char *ptr = SvPVutf8((SV*)rv, len);
- Cfish_CB_Mimic_Str(self->pattern, ptr, len);
+ Cfish_CB_Mimic_Str(ivars->pattern, ptr, len);
SvREFCNT_dec(rv);
}
@@ -125,8 +128,9 @@ lucy_RegexTokenizer_set_token_re(lucy_RegexTokenizer *self, void *token_re) {
void
lucy_RegexTokenizer_destroy(lucy_RegexTokenizer *self) {
- CFISH_DECREF(self->pattern);
- ReREFCNT_dec(((REGEXP*)self->token_re));
+ lucy_RegexTokenizerIVARS *const ivars = lucy_RegexTokenizer_IVARS(self);
+ CFISH_DECREF(ivars->pattern);
+ ReREFCNT_dec(((REGEXP*)ivars->token_re));
CFISH_SUPER_DESTROY(self, LUCY_REGEXTOKENIZER);
}
@@ -134,13 +138,14 @@ void
lucy_RegexTokenizer_tokenize_str(lucy_RegexTokenizer *self,
const char *string, size_t string_len,
lucy_Inversion *inversion) {
+ lucy_RegexTokenizerIVARS *const ivars = lucy_RegexTokenizer_IVARS(self);
uint32_t num_code_points = 0;
SV *wrapper = sv_newmortal();
#if (PERL_VERSION > 10)
- REGEXP *rx = (REGEXP*)self->token_re;
+ REGEXP *rx = (REGEXP*)ivars->token_re;
regexp *rx_struct = (regexp*)SvANY(rx);
#else
- REGEXP *rx = (REGEXP*)self->token_re;
+ REGEXP *rx = (REGEXP*)ivars->token_re;
regexp *rx_struct = rx;
#endif
char *string_beg = (char*)string;
http://git-wip-us.apache.org/repos/asf/lucy/blob/811358d9/perl/xs/Lucy/Document/Doc.c
----------------------------------------------------------------------
diff --git a/perl/xs/Lucy/Document/Doc.c b/perl/xs/Lucy/Document/Doc.c
index c20c2b0..220a7bf 100644
--- a/perl/xs/Lucy/Document/Doc.c
+++ b/perl/xs/Lucy/Document/Doc.c
@@ -23,32 +23,36 @@
lucy_Doc*
lucy_Doc_init(lucy_Doc *self, void *fields, int32_t doc_id) {
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
// Assign.
if (fields) {
if (SvTYPE((SV*)fields) != SVt_PVHV) { THROW(CFISH_ERR, "Not a hash"); }
- self->fields = SvREFCNT_inc((SV*)fields);
+ ivars->fields = SvREFCNT_inc((SV*)fields);
}
else {
- self->fields = newHV();
+ ivars->fields = newHV();
}
- self->doc_id = doc_id;
+ ivars->doc_id = doc_id;
return self;
}
void
lucy_Doc_set_fields(lucy_Doc *self, void *fields) {
- if (self->fields) { SvREFCNT_dec((SV*)self->fields); }
- self->fields = SvREFCNT_inc((SV*)fields);
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
+ if (ivars->fields) { SvREFCNT_dec((SV*)ivars->fields); }
+ ivars->fields = SvREFCNT_inc((SV*)fields);
}
uint32_t
lucy_Doc_get_size(lucy_Doc *self) {
- return self->fields ? HvKEYS((HV*)self->fields) : 0;
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
+ return ivars->fields ? HvKEYS((HV*)ivars->fields) : 0;
}
void
lucy_Doc_store(lucy_Doc *self, const cfish_CharBuf *field, cfish_Obj *value) {
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
char *key = (char*)Cfish_CB_Get_Ptr8(field);
size_t key_size = Cfish_CB_Get_Size(field);
SV *key_sv = newSVpvn(key, key_size);
@@ -58,19 +62,20 @@ lucy_Doc_store(lucy_Doc *self, const cfish_CharBuf *field, cfish_Obj *value) {
? XSBind_cb_to_sv((cfish_CharBuf*)value)
: (SV*)Cfish_Obj_To_Host(value);
SvUTF8_on(key_sv);
- (void)hv_store_ent((HV*)self->fields, key_sv, val_sv, 0);
+ (void)hv_store_ent((HV*)ivars->fields, key_sv, val_sv, 0);
// TODO: make this a thread-local instead of creating it every time?
SvREFCNT_dec(key_sv);
}
static SV*
S_nfreeze_fields(lucy_Doc *self) {
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
dSP;
ENTER;
SAVETMPS;
EXTEND(SP, 1);
PUSHMARK(SP);
- mPUSHs((SV*)newRV_inc((SV*)self->fields));
+ mPUSHs((SV*)newRV_inc((SV*)ivars->fields));
PUTBACK;
call_pv("Storable::nfreeze", G_SCALAR);
SPAGAIN;
@@ -84,7 +89,8 @@ S_nfreeze_fields(lucy_Doc *self) {
void
lucy_Doc_serialize(lucy_Doc *self, lucy_OutStream *outstream) {
- Lucy_OutStream_Write_C32(outstream, self->doc_id);
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
+ Lucy_OutStream_Write_C32(outstream, ivars->doc_id);
SV *frozen = S_nfreeze_fields(self);
STRLEN len;
char *buf = SvPV(frozen, len);
@@ -138,8 +144,9 @@ lucy_Doc_deserialize(lucy_Doc *self, lucy_InStream *instream) {
cfish_Obj*
lucy_Doc_extract(lucy_Doc *self, cfish_CharBuf *field,
cfish_ViewCharBuf *target) {
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
cfish_Obj *retval = NULL;
- SV **sv_ptr = hv_fetch((HV*)self->fields, (char*)Cfish_CB_Get_Ptr8(field),
+ SV **sv_ptr = hv_fetch((HV*)ivars->fields, (char*)Cfish_CB_Get_Ptr8(field),
Cfish_CB_Get_Size(field), 0);
if (sv_ptr && XSBind_sv_defined(*sv_ptr)) {
@@ -170,13 +177,14 @@ lucy_Doc_to_host(lucy_Doc *self) {
cfish_Hash*
lucy_Doc_dump(lucy_Doc *self) {
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
cfish_Hash *dump = cfish_Hash_new(0);
Cfish_Hash_Store_Str(dump, "_class", 6,
(cfish_Obj*)Cfish_CB_Clone(Lucy_Doc_Get_Class_Name(self)));
Cfish_Hash_Store_Str(dump, "doc_id", 7,
- (cfish_Obj*)cfish_CB_newf("%i32", self->doc_id));
+ (cfish_Obj*)cfish_CB_newf("%i32", ivars->doc_id));
Cfish_Hash_Store_Str(dump, "fields", 6,
- XSBind_perl_to_cfish((SV*)self->fields));
+ XSBind_perl_to_cfish((SV*)ivars->fields));
return dump;
}
@@ -197,8 +205,9 @@ lucy_Doc_load(lucy_Doc *self, cfish_Obj *dump) {
SV *fields_sv = XSBind_cfish_to_perl((cfish_Obj*)fields);
CHY_UNUSED_VAR(self);
- loaded->doc_id = (int32_t)Cfish_Obj_To_I64(doc_id);
- loaded->fields = SvREFCNT_inc(SvRV(fields_sv));
+ lucy_DocIVARS *const loaded_ivars = lucy_Doc_IVARS(loaded);
+ loaded_ivars->doc_id = (int32_t)Cfish_Obj_To_I64(doc_id);
+ loaded_ivars->fields = SvREFCNT_inc(SvRV(fields_sv));
SvREFCNT_dec(fields_sv);
return loaded;
@@ -206,21 +215,19 @@ lucy_Doc_load(lucy_Doc *self, cfish_Obj *dump) {
bool
lucy_Doc_equals(lucy_Doc *self, cfish_Obj *other) {
- lucy_Doc *twin = (lucy_Doc*)other;
- HV *my_fields;
- HV *other_fields;
- I32 num_fields;
-
- if (twin == self) { return true; }
+ if ((lucy_Doc*)other == self) { return true; }
if (!Cfish_Obj_Is_A(other, LUCY_DOC)) { return false; }
- if (!self->doc_id == twin->doc_id) { return false; }
- if (!!self->fields ^ !!twin->fields) { return false; }
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
+ lucy_DocIVARS *const ovars = lucy_Doc_IVARS((lucy_Doc*)other);
+
+ if (!ivars->doc_id == ovars->doc_id) { return false; }
+ if (!!ivars->fields ^ !!ovars->fields) { return false; }
// Verify fields. Don't allow any deep data structures.
- my_fields = (HV*)self->fields;
- other_fields = (HV*)twin->fields;
+ HV *my_fields = (HV*)ivars->fields;
+ HV *other_fields = (HV*)ovars->fields;
if (HvKEYS(my_fields) != HvKEYS(other_fields)) { return false; }
- num_fields = hv_iterinit(my_fields);
+ I32 num_fields = hv_iterinit(my_fields);
while (num_fields--) {
HE *my_entry = hv_iternext(my_fields);
SV *my_val_sv = HeVAL(my_entry);
@@ -236,7 +243,8 @@ lucy_Doc_equals(lucy_Doc *self, cfish_Obj *other) {
void
lucy_Doc_destroy(lucy_Doc *self) {
- if (self->fields) { SvREFCNT_dec((SV*)self->fields); }
+ lucy_DocIVARS *const ivars = lucy_Doc_IVARS(self);
+ if (ivars->fields) { SvREFCNT_dec((SV*)ivars->fields); }
CFISH_SUPER_DESTROY(self, LUCY_DOC);
}
http://git-wip-us.apache.org/repos/asf/lucy/blob/811358d9/perl/xs/Lucy/Index/DocReader.c
----------------------------------------------------------------------
diff --git a/perl/xs/Lucy/Index/DocReader.c b/perl/xs/Lucy/Index/DocReader.c
index 0c7b2ac..d4e27ae 100644
--- a/perl/xs/Lucy/Index/DocReader.c
+++ b/perl/xs/Lucy/Index/DocReader.c
@@ -29,9 +29,10 @@
lucy_HitDoc*
lucy_DefDocReader_fetch_doc(lucy_DefaultDocReader *self, int32_t doc_id) {
- lucy_Schema *const schema = self->schema;
- lucy_InStream *const dat_in = self->dat_in;
- lucy_InStream *const ix_in = self->ix_in;
+ lucy_DefaultDocReaderIVARS *const ivars = lucy_DefDocReader_IVARS(self);
+ lucy_Schema *const schema = ivars->schema;
+ lucy_InStream *const dat_in = ivars->dat_in;
+ lucy_InStream *const ix_in = ivars->ix_in;
HV *fields = newHV();
int64_t start;
uint32_t num_fields;
http://git-wip-us.apache.org/repos/asf/lucy/blob/811358d9/perl/xs/Lucy/Index/Inverter.c
----------------------------------------------------------------------
diff --git a/perl/xs/Lucy/Index/Inverter.c b/perl/xs/Lucy/Index/Inverter.c
index 7f0bc95..d3ad2a8 100644
--- a/perl/xs/Lucy/Index/Inverter.c
+++ b/perl/xs/Lucy/Index/Inverter.c
@@ -30,7 +30,8 @@
static lucy_InverterEntry*
S_fetch_entry(lucy_Inverter *self, HE *hash_entry) {
- lucy_Schema *const schema = self->schema;
+ lucy_InverterIVARS *const ivars = lucy_Inverter_IVARS(self);
+ lucy_Schema *const schema = ivars->schema;
char *key;
STRLEN key_len;
STRLEN he_key_len = HeKLEN(hash_entry);
@@ -51,13 +52,13 @@ S_fetch_entry(lucy_Inverter *self, HE *hash_entry) {
cfish_ZombieCharBuf *field = CFISH_ZCB_WRAP_STR(key, key_len);
int32_t field_num
- = Lucy_Seg_Field_Num(self->segment, (cfish_CharBuf*)field);
+ = Lucy_Seg_Field_Num(ivars->segment, (cfish_CharBuf*)field);
if (!field_num) {
// This field seems not to be in the segment yet. Try to find it in
// the Schema.
if (Lucy_Schema_Fetch_Type(schema, (cfish_CharBuf*)field)) {
// The field is in the Schema. Get a field num from the Segment.
- field_num = Lucy_Seg_Add_Field(self->segment,
+ field_num = Lucy_Seg_Add_Field(ivars->segment,
(cfish_CharBuf*)field);
}
else {
@@ -68,10 +69,10 @@ S_fetch_entry(lucy_Inverter *self, HE *hash_entry) {
}
lucy_InverterEntry *entry
- = (lucy_InverterEntry*)Cfish_VA_Fetch(self->entry_pool, field_num);
+ = (lucy_InverterEntry*)Cfish_VA_Fetch(ivars->entry_pool, field_num);
if (!entry) {
entry = lucy_InvEntry_new(schema, (cfish_CharBuf*)field, field_num);
- Cfish_VA_Store(self->entry_pool, field_num, (cfish_Obj*)entry);
+ Cfish_VA_Store(ivars->entry_pool, field_num, (cfish_Obj*)entry);
}
return entry;
}
@@ -89,7 +90,9 @@ lucy_Inverter_invert_doc(lucy_Inverter *self, lucy_Doc *doc) {
HE *hash_entry = hv_iternext(fields);
lucy_InverterEntry *inv_entry = S_fetch_entry(self, hash_entry);
SV *value_sv = HeVAL(hash_entry);
- lucy_FieldType *type = inv_entry->type;
+ lucy_InverterEntryIVARS *const entry_ivars
+ = lucy_InvEntry_IVARS(inv_entry);
+ lucy_FieldType *type = entry_ivars->type;
// Get the field value, forcing text fields to UTF-8.
switch (Lucy_FType_Primitive_ID(type) & lucy_FType_PRIMITIVE_ID_MASK) {
@@ -97,7 +100,7 @@ lucy_Inverter_invert_doc(lucy_Inverter *self, lucy_Doc *doc) {
STRLEN val_len;
char *val_ptr = SvPVutf8(value_sv, val_len);
cfish_ViewCharBuf *value
- = (cfish_ViewCharBuf*)inv_entry->value;
+ = (cfish_ViewCharBuf*)entry_ivars->value;
Cfish_ViewCB_Assign_Str(value, val_ptr, val_len);
break;
}
@@ -105,17 +108,17 @@ lucy_Inverter_invert_doc(lucy_Inverter *self, lucy_Doc *doc) {
STRLEN val_len;
char *val_ptr = SvPV(value_sv, val_len);
cfish_ViewByteBuf *value
- = (cfish_ViewByteBuf*)inv_entry->value;
+ = (cfish_ViewByteBuf*)entry_ivars->value;
Cfish_ViewBB_Assign_Bytes(value, val_ptr, val_len);
break;
}
case lucy_FType_INT32: {
- cfish_Integer32* value = (cfish_Integer32*)inv_entry->value;
+ cfish_Integer32* value = (cfish_Integer32*)entry_ivars->value;
Cfish_Int32_Set_Value(value, SvIV(value_sv));
break;
}
case lucy_FType_INT64: {
- cfish_Integer64* value = (cfish_Integer64*)inv_entry->value;
+ cfish_Integer64* value = (cfish_Integer64*)entry_ivars->value;
int64_t val = sizeof(IV) == 8
? SvIV(value_sv)
: (int64_t)SvNV(value_sv); // lossy
@@ -123,12 +126,12 @@ lucy_Inverter_invert_doc(lucy_Inverter *self, lucy_Doc *doc) {
break;
}
case lucy_FType_FLOAT32: {
- cfish_Float32* value = (cfish_Float32*)inv_entry->value;
+ cfish_Float32* value = (cfish_Float32*)entry_ivars->value;
Cfish_Float32_Set_Value(value, (float)SvNV(value_sv));
break;
}
case lucy_FType_FLOAT64: {
- cfish_Float64* value = (cfish_Float64*)inv_entry->value;
+ cfish_Float64* value = (cfish_Float64*)entry_ivars->value;
Cfish_Float64_Set_Value(value, SvNV(value_sv));
break;
}