ICU-21035 add & use CharString::extract(dest buffer)
See #1253
diff --git a/icu4c/source/common/charstr.cpp b/icu4c/source/common/charstr.cpp
index 4878c73..318a185 100644
--- a/icu4c/source/common/charstr.cpp
+++ b/icu4c/source/common/charstr.cpp
@@ -20,6 +20,7 @@
#include "cmemory.h"
#include "cstring.h"
#include "uinvchar.h"
+#include "ustr_imp.h"
U_NAMESPACE_BEGIN
@@ -46,6 +47,19 @@
return p;
}
+int32_t CharString::extract(char *dest, int32_t capacity, UErrorCode &errorCode) const {
+ if (U_FAILURE(errorCode)) { return len; }
+ if (capacity < 0 || (capacity > 0 && dest == nullptr)) {
+ errorCode = U_ILLEGAL_ARGUMENT_ERROR;
+ return len;
+ }
+ const char *src = buffer.getAlias();
+ if (0 < len && len <= capacity && src != dest) {
+ uprv_memcpy(dest, src, len);
+ }
+ return u_terminateChars(dest, capacity, len, &errorCode);
+}
+
CharString &CharString::copyFrom(const CharString &s, UErrorCode &errorCode) {
if(U_SUCCESS(errorCode) && this!=&s && ensureCapacity(s.len+1, 0, errorCode)) {
len=s.len;
diff --git a/icu4c/source/common/charstr.h b/icu4c/source/common/charstr.h
index 9400fbb..6619faa 100644
--- a/icu4c/source/common/charstr.h
+++ b/icu4c/source/common/charstr.h
@@ -87,6 +87,22 @@
* The caller must uprv_free() the result.
*/
char *cloneData(UErrorCode &errorCode) const;
+ /**
+ * Copies the contents of the string into dest.
+ * Checks if there is enough space in dest, extracts the entire string if possible,
+ * and NUL-terminates dest if possible.
+ *
+ * If the string fits into dest but cannot be NUL-terminated (length()==capacity),
+ * then the error code is set to U_STRING_NOT_TERMINATED_WARNING.
+ * If the string itself does not fit into dest (length()>capacity),
+ * then the error code is set to U_BUFFER_OVERFLOW_ERROR.
+ *
+ * @param dest Destination string buffer.
+ * @param capacity Size of the dest buffer (number of chars).
+ * @param errorCode ICU error code.
+ * @return length()
+ */
+ int32_t extract(char *dest, int32_t capacity, UErrorCode &errorCode) const;
bool operator==(StringPiece other) const {
return len == other.length() && (len == 0 || uprv_memcmp(data(), other.data(), len) == 0);
diff --git a/icu4c/source/common/icuplug.cpp b/icu4c/source/common/icuplug.cpp
index c6439cc..303bd44 100644
--- a/icu4c/source/common/icuplug.cpp
+++ b/icu4c/source/common/icuplug.cpp
@@ -782,8 +782,8 @@
/* plugin_file is not used for processing - it is only used
so that uplug_getPluginFile() works (i.e. icuinfo)
*/
- uprv_strncpy(plugin_file, pluginFile.data(), sizeof(plugin_file));
-
+ pluginFile.extract(plugin_file, sizeof(plugin_file), *status);
+
#if UPLUG_TRACE
DBG((stderr, "pluginfile= %s len %d/%d\n", plugin_file, (int)strlen(plugin_file), (int)sizeof(plugin_file)));
#endif
diff --git a/icu4c/source/common/loclikely.cpp b/icu4c/source/common/loclikely.cpp
index ea0d98c..e87b06e 100644
--- a/icu4c/source/common/loclikely.cpp
+++ b/icu4c/source/common/loclikely.cpp
@@ -464,18 +464,7 @@
goto error;
}
- {
- icu::CharString result = ulocimp_getLanguage(position, &position, *err);
- if (U_FAILURE(*err)) {
- goto error;
- }
-
- subtagLength = result.length();
- if (subtagLength <= *langLength) {
- uprv_memcpy(lang, result.data(), subtagLength);
- }
- u_terminateChars(lang, *langLength, subtagLength, err);
- }
+ subtagLength = ulocimp_getLanguage(position, &position, *err).extract(lang, *langLength, *err);
/*
* Note that we explicit consider U_STRING_NOT_TERMINATED_WARNING
diff --git a/icu4c/source/common/uloc.cpp b/icu4c/source/common/uloc.cpp
index 94cbaab..45fab01 100644
--- a/icu4c/source/common/uloc.cpp
+++ b/icu4c/source/common/uloc.cpp
@@ -50,9 +50,6 @@
#include "uassert.h"
#include "charstr.h"
-#include <algorithm>
-#include <stdio.h> /* for sprintf */
-
U_NAMESPACE_USE
/* ### Declarations **************************************************/
@@ -892,7 +889,6 @@
char* startSearchHere = NULL;
char* keywordStart = NULL;
CharString updatedKeysAndValues;
- int32_t updatedKeysAndValuesLen;
UBool handledInputKeyAndValue = FALSE;
char keyValuePrefix = '@';
@@ -1072,18 +1068,10 @@
return bufLen;
}
- updatedKeysAndValuesLen = updatedKeysAndValues.length();
- /* needLen = length of the part before '@' + length of updated key-value part including '@' */
- needLen = (int32_t)(startSearchHere - buffer) + updatedKeysAndValuesLen;
- if(needLen >= bufferCapacity) {
- *status = U_BUFFER_OVERFLOW_ERROR;
- return needLen; /* no change */
- }
- if (updatedKeysAndValuesLen > 0) {
- uprv_strncpy(startSearchHere, updatedKeysAndValues.data(), updatedKeysAndValuesLen);
- }
- buffer[needLen]=0;
- return needLen;
+ // needLen = length of the part before '@'
+ needLen = (int32_t)(startSearchHere - buffer);
+ return needLen + updatedKeysAndValues.extract(
+ startSearchHere, bufferCapacity - needLen, *status);
}
/* ### ID parsing implementation **************************************************/
@@ -1232,13 +1220,7 @@
char *script, int32_t scriptCapacity,
const char **pEnd) {
ErrorCode status;
- CharString result = ulocimp_getScript(localeID, pEnd, status);
- if (status.isFailure()) {
- return 0;
- }
- int32_t reslen = result.length();
- uprv_memcpy(script, result.data(), std::min(reslen, scriptCapacity));
- return reslen;
+ return ulocimp_getScript(localeID, pEnd, status).extract(script, scriptCapacity, status);
}
static CharString
@@ -1281,13 +1263,7 @@
char *country, int32_t countryCapacity,
const char **pEnd) {
ErrorCode status;
- CharString result = ulocimp_getCountry(localeID, pEnd, status);
- if (status.isFailure()) {
- return 0;
- }
- int32_t reslen = result.length();
- uprv_memcpy(country, result.data(), std::min(reslen, countryCapacity));
- return reslen;
+ return ulocimp_getCountry(localeID, pEnd, status).extract(country, countryCapacity, status);
}
/**
@@ -1744,16 +1720,7 @@
localeID=uloc_getDefault();
}
- CharString result = ulocimp_getLanguage(localeID, NULL, *err);
- if (U_FAILURE(*err)) {
- return 0;
- }
-
- int32_t reslen = result.length();
- if (reslen <= languageCapacity) {
- uprv_memcpy(language, result.data(), reslen);
- }
- return u_terminateChars(language, languageCapacity, reslen, err);
+ return ulocimp_getLanguage(localeID, NULL, *err).extract(language, languageCapacity, *err);
}
U_CAPI int32_t U_EXPORT2
diff --git a/icu4c/source/common/uloc_tag.cpp b/icu4c/source/common/uloc_tag.cpp
index 5eed02c..8faac32 100644
--- a/icu4c/source/common/uloc_tag.cpp
+++ b/icu4c/source/common/uloc_tag.cpp
@@ -1413,7 +1413,7 @@
break;
}
- uprv_strcpy(pExtBuf, bcpValue);
+ buf.extract(pExtBuf, resultCapacity, tmpStatus);
T_CString_toLowerCase(pExtBuf);
extBuf->append(pExtBuf, bcpValueLen, tmpStatus);
diff --git a/icu4c/source/i18n/rulebasedcollator.cpp b/icu4c/source/i18n/rulebasedcollator.cpp
index 60acf17..917482d 100644
--- a/icu4c/source/i18n/rulebasedcollator.cpp
+++ b/icu4c/source/i18n/rulebasedcollator.cpp
@@ -1600,10 +1600,7 @@
appendSubtag(result, 'Z', subtag, length, errorCode);
if(U_FAILURE(errorCode)) { return 0; }
- if(result.length() <= capacity) {
- uprv_memcpy(buffer, result.data(), result.length());
- }
- return u_terminateChars(buffer, capacity, result.length(), &errorCode);
+ return result.extract(buffer, capacity, errorCode);
}
UBool
diff --git a/icu4c/source/i18n/ucol_sit.cpp b/icu4c/source/i18n/ucol_sit.cpp
index 92f332d..4dc81ae 100644
--- a/icu4c/source/i18n/ucol_sit.cpp
+++ b/icu4c/source/i18n/ucol_sit.cpp
@@ -372,10 +372,7 @@
}
len += s->entries[i].length();
} else {
- len += s->entries[i].length();
- if(len < capacity) {
- uprv_strncat(destination,s->entries[i].data(), s->entries[i].length());
- }
+ len += s->entries[i].extract(destination + len, capacity - len, *status);
}
}
}
diff --git a/icu4c/source/test/cintltst/cloctst.c b/icu4c/source/test/cintltst/cloctst.c
index 29f4c82..f87e518 100644
--- a/icu4c/source/test/cintltst/cloctst.c
+++ b/icu4c/source/test/cintltst/cloctst.c
@@ -2233,7 +2233,12 @@
strcpy(buffer,kwSetTestCases[i].l);
status = U_ZERO_ERROR;
res = uloc_setKeywordValue(kwSetTestCases[i].k, kwSetTestCases[i].v, buffer, blen, &status);
- if(status != U_BUFFER_OVERFLOW_ERROR) {
+ if(res == blen) {
+ if(status != U_STRING_NOT_TERMINATED_WARNING) {
+ log_err("expected not terminated warning on buffer %d got %s, len %d (%s + [%s=%s])\n", blen, u_errorName(status), res, kwSetTestCases[i].l, kwSetTestCases[i].k, kwSetTestCases[i].v);
+ return;
+ }
+ } else if(status != U_BUFFER_OVERFLOW_ERROR) {
log_err("expected buffer overflow on buffer %d got %s, len %d (%s + [%s=%s])\n", blen, u_errorName(status), res, kwSetTestCases[i].l, kwSetTestCases[i].k, kwSetTestCases[i].v);
return;
}
diff --git a/icu4c/source/test/intltest/strtest.cpp b/icu4c/source/test/intltest/strtest.cpp
index 1185d31..54b5a19 100644
--- a/icu4c/source/test/intltest/strtest.cpp
+++ b/icu4c/source/test/intltest/strtest.cpp
@@ -810,6 +810,27 @@
"Long string over 40 characters to trigger heap allocation",
s3.data());
}
+
+ {
+ // extract()
+ errorCode.reset();
+ CharString s("abc", errorCode);
+ char buffer[10];
+
+ s.extract(buffer, 10, errorCode);
+ assertEquals("abc.extract(10) success", U_ZERO_ERROR, errorCode.get());
+ assertEquals("abc.extract(10) output", "abc", buffer);
+
+ strcpy(buffer, "012345");
+ s.extract(buffer, 3, errorCode);
+ assertEquals("abc.extract(3) not terminated",
+ U_STRING_NOT_TERMINATED_WARNING, errorCode.reset());
+ assertEquals("abc.extract(3) output", "abc345", buffer);
+
+ strcpy(buffer, "012345");
+ s.extract(buffer, 2, errorCode);
+ assertEquals("abc.extract(2) overflow", U_BUFFER_OVERFLOW_ERROR, errorCode.reset());
+ }
}
void
diff --git a/icu4c/source/tools/toolutil/pkg_genc.cpp b/icu4c/source/tools/toolutil/pkg_genc.cpp
index 8251128..17347ba 100644
--- a/icu4c/source/tools/toolutil/pkg_genc.cpp
+++ b/icu4c/source/tools/toolutil/pkg_genc.cpp
@@ -738,8 +738,8 @@
exit(U_ILLEGAL_ARGUMENT_ERROR);
}
- uprv_strcpy(outFilename, outFilenameBuilder.data());
- uprv_strcpy(entryName, entryNameBuilder.data());
+ outFilenameBuilder.extract(outFilename, outFilenameCapacity, status);
+ entryNameBuilder.extract(entryName, entryNameCapacity, status);
}
#ifdef CAN_GENERATE_OBJECTS
