diff options
Diffstat (limited to 'src/libutil')
-rw-r--r-- | src/libutil/archive.cc | 10 | ||||
-rw-r--r-- | src/libutil/archive.hh | 6 | ||||
-rw-r--r-- | src/libutil/hash.cc | 95 | ||||
-rw-r--r-- | src/libutil/hash.hh | 30 | ||||
-rw-r--r-- | src/libutil/serialise.hh | 16 | ||||
-rw-r--r-- | src/libutil/split.hh | 33 | ||||
-rw-r--r-- | src/libutil/topo-sort.hh | 40 | ||||
-rw-r--r-- | src/libutil/util.cc | 11 | ||||
-rw-r--r-- | src/libutil/util.hh | 2 |
9 files changed, 181 insertions, 62 deletions
diff --git a/src/libutil/archive.cc b/src/libutil/archive.cc index 51c88537e..ce7cf9754 100644 --- a/src/libutil/archive.cc +++ b/src/libutil/archive.cc @@ -150,17 +150,17 @@ static void skipGeneric(Source & source) static void parseContents(ParseSink & sink, Source & source, const Path & path) { - unsigned long long size = readLongLong(source); + uint64_t size = readLongLong(source); sink.preallocateContents(size); - unsigned long long left = size; + uint64_t left = size; std::vector<unsigned char> buf(65536); while (left) { checkInterrupt(); auto n = buf.size(); - if ((unsigned long long)n > left) n = left; + if ((uint64_t)n > left) n = left; source(buf.data(), n); sink.receiveContents(buf.data(), n); left -= n; @@ -323,7 +323,7 @@ struct RestoreSink : ParseSink throw SysError("fchmod"); } - void preallocateContents(unsigned long long len) + void preallocateContents(uint64_t len) { #if HAVE_POSIX_FALLOCATE if (len) { @@ -338,7 +338,7 @@ struct RestoreSink : ParseSink #endif } - void receiveContents(unsigned char * data, unsigned int len) + void receiveContents(unsigned char * data, size_t len) { writeFull(fd.get(), data, len); } diff --git a/src/libutil/archive.hh b/src/libutil/archive.hh index 57780d16a..5665732d2 100644 --- a/src/libutil/archive.hh +++ b/src/libutil/archive.hh @@ -57,8 +57,8 @@ struct ParseSink virtual void createRegularFile(const Path & path) { }; virtual void isExecutable() { }; - virtual void preallocateContents(unsigned long long size) { }; - virtual void receiveContents(unsigned char * data, unsigned int len) { }; + virtual void preallocateContents(uint64_t size) { }; + virtual void receiveContents(unsigned char * data, size_t len) { }; virtual void createSymlink(const Path & path, const string & target) { }; }; @@ -77,7 +77,7 @@ struct RetrieveRegularNARSink : ParseSink regular = false; } - void receiveContents(unsigned char * data, unsigned int len) + void receiveContents(unsigned char * data, size_t len) { sink(data, len); } diff --git a/src/libutil/hash.cc b/src/libutil/hash.cc index 2b0390da4..dfb3668f1 100644 --- a/src/libutil/hash.cc +++ b/src/libutil/hash.cc @@ -7,6 +7,7 @@ #include "args.hh" #include "hash.hh" #include "archive.hh" +#include "split.hh" #include "util.hh" #include <sys/types.h> @@ -15,6 +16,7 @@ namespace nix { + static size_t regularHashSize(HashType type) { switch (type) { case htMD5: return md5HashSize; @@ -25,10 +27,11 @@ static size_t regularHashSize(HashType type) { abort(); } + std::set<std::string> hashTypes = { "md5", "sha1", "sha256", "sha512" }; -void Hash::init() +Hash::Hash(HashType type) : type(type) { hashSize = regularHashSize(type); assert(hashSize <= maxHashSize); @@ -133,57 +136,89 @@ std::string Hash::to_string(Base base, bool includeType) const return s; } -Hash::Hash(std::string_view s, HashType type) : Hash(s, std::optional { type }) { } -Hash::Hash(std::string_view s) : Hash(s, std::optional<HashType>{}) { } - -Hash::Hash(std::string_view original, std::optional<HashType> optType) -{ +Hash Hash::parseSRI(std::string_view original) { auto rest = original; - size_t pos = 0; + // Parse the has type before the separater, if there was one. + auto hashRaw = splitPrefixTo(rest, '-'); + if (!hashRaw) + throw BadHash("hash '%s' is not SRI", original); + HashType parsedType = parseHashType(*hashRaw); + + return Hash(rest, parsedType, true); +} + +// Mutates the string to eliminate the prefixes when found +static std::pair<std::optional<HashType>, bool> getParsedTypeAndSRI(std::string_view & rest) { bool isSRI = false; // Parse the has type before the separater, if there was one. std::optional<HashType> optParsedType; { - auto sep = rest.find(':'); - if (sep == std::string_view::npos) { - sep = rest.find('-'); - if (sep != std::string_view::npos) + auto hashRaw = splitPrefixTo(rest, ':'); + + if (!hashRaw) { + hashRaw = splitPrefixTo(rest, '-'); + if (hashRaw) isSRI = true; } - if (sep != std::string_view::npos) { - auto hashRaw = rest.substr(0, sep); - optParsedType = parseHashType(hashRaw); - rest = rest.substr(sep + 1); - } + if (hashRaw) + optParsedType = parseHashType(*hashRaw); } + return {optParsedType, isSRI}; +} + +Hash Hash::parseAnyPrefixed(std::string_view original) +{ + auto rest = original; + auto [optParsedType, isSRI] = getParsedTypeAndSRI(rest); + // Either the string or user must provide the type, if they both do they // must agree. - if (!optParsedType && !optType) { + if (!optParsedType) + throw BadHash("hash '%s' does not include a type", rest); + + return Hash(rest, *optParsedType, isSRI); +} + +Hash Hash::parseAny(std::string_view original, std::optional<HashType> optType) +{ + auto rest = original; + auto [optParsedType, isSRI] = getParsedTypeAndSRI(rest); + + // Either the string or user must provide the type, if they both do they + // must agree. + if (!optParsedType && !optType) throw BadHash("hash '%s' does not include a type, nor is the type otherwise known from context.", rest); - } else { - this->type = optParsedType ? *optParsedType : *optType; - if (optParsedType && optType && *optParsedType != *optType) - throw BadHash("hash '%s' should have type '%s'", original, printHashType(*optType)); - } + else if (optParsedType && optType && *optParsedType != *optType) + throw BadHash("hash '%s' should have type '%s'", original, printHashType(*optType)); - init(); + HashType hashType = optParsedType ? *optParsedType : *optType; + return Hash(rest, hashType, isSRI); +} +Hash Hash::parseNonSRIUnprefixed(std::string_view s, HashType type) +{ + return Hash(s, type, false); +} + +Hash::Hash(std::string_view rest, HashType type, bool isSRI) + : Hash(type) +{ if (!isSRI && rest.size() == base16Len()) { auto parseHexDigit = [&](char c) { if (c >= '0' && c <= '9') return c - '0'; if (c >= 'A' && c <= 'F') return c - 'A' + 10; if (c >= 'a' && c <= 'f') return c - 'a' + 10; - throw BadHash("invalid base-16 hash '%s'", original); + throw BadHash("invalid base-16 hash '%s'", rest); }; for (unsigned int i = 0; i < hashSize; i++) { hash[i] = - parseHexDigit(rest[pos + i * 2]) << 4 - | parseHexDigit(rest[pos + i * 2 + 1]); + parseHexDigit(rest[i * 2]) << 4 + | parseHexDigit(rest[i * 2 + 1]); } } @@ -195,7 +230,7 @@ Hash::Hash(std::string_view original, std::optional<HashType> optType) for (digit = 0; digit < base32Chars.size(); ++digit) /* !!! slow */ if (base32Chars[digit] == c) break; if (digit >= 32) - throw BadHash("invalid base-32 hash '%s'", original); + throw BadHash("invalid base-32 hash '%s'", rest); unsigned int b = n * 5; unsigned int i = b / 8; unsigned int j = b % 8; @@ -205,7 +240,7 @@ Hash::Hash(std::string_view original, std::optional<HashType> optType) hash[i + 1] |= digit >> (8 - j); } else { if (digit >> (8 - j)) - throw BadHash("invalid base-32 hash '%s'", original); + throw BadHash("invalid base-32 hash '%s'", rest); } } } @@ -213,7 +248,7 @@ Hash::Hash(std::string_view original, std::optional<HashType> optType) else if (isSRI || rest.size() == base64Len()) { auto d = base64Decode(rest); if (d.size() != hashSize) - throw BadHash("invalid %s hash '%s'", isSRI ? "SRI" : "base-64", original); + throw BadHash("invalid %s hash '%s'", isSRI ? "SRI" : "base-64", rest); assert(hashSize); memcpy(hash, d.data(), hashSize); } @@ -231,7 +266,7 @@ Hash newHashAllowEmpty(std::string hashStr, std::optional<HashType> ht) warn("found empty hash, assuming '%s'", h.to_string(SRI, true)); return h; } else - return Hash(hashStr, ht); + return Hash::parseAny(hashStr, ht); } diff --git a/src/libutil/hash.hh b/src/libutil/hash.hh index 98ee1bed0..00ce7bb6f 100644 --- a/src/libutil/hash.hh +++ b/src/libutil/hash.hh @@ -34,21 +34,31 @@ struct Hash HashType type; /* Create a zero-filled hash object. */ - Hash(HashType type) : type(type) { init(); }; + Hash(HashType type); - /* Initialize the hash from a string representation, in the format + /* Parse the hash from a string representation in the format "[<type>:]<base16|base32|base64>" or "<type>-<base64>" (a Subresource Integrity hash expression). If the 'type' argument is not present, then the hash type must be specified in the string. */ - Hash(std::string_view s, std::optional<HashType> type); - // type must be provided - Hash(std::string_view s, HashType type); - // hash type must be part of string - Hash(std::string_view s); + static Hash parseAny(std::string_view s, std::optional<HashType> type); - void init(); + /* Parse a hash from a string representation like the above, except the + type prefix is mandatory is there is no separate arguement. */ + static Hash parseAnyPrefixed(std::string_view s); + /* Parse a plain hash that musst not have any prefix indicating the type. + The type is passed in to disambiguate. */ + static Hash parseNonSRIUnprefixed(std::string_view s, HashType type); + + static Hash parseSRI(std::string_view original); + +private: + /* The type must be provided, the string view must not include <type> + prefix. `isSRI` helps disambigate the various base-* encodings. */ + Hash(std::string_view s, HashType type, bool isSRI); + +public: /* Check whether a hash is set. */ operator bool () const { return (bool) type; } @@ -111,7 +121,7 @@ Hash hashFile(HashType ht, const Path & path); /* Compute the hash of the given path. The hash is defined as (essentially) hashString(ht, dumpPath(path)). */ -typedef std::pair<Hash, unsigned long long> HashResult; +typedef std::pair<Hash, uint64_t> HashResult; HashResult hashPath(HashType ht, const Path & path, PathFilter & filter = defaultPathFilter); @@ -141,7 +151,7 @@ class HashSink : public BufferedSink, public AbstractHashSink private: HashType ht; Ctx * ctx; - unsigned long long bytes; + uint64_t bytes; public: HashSink(HashType ht); diff --git a/src/libutil/serialise.hh b/src/libutil/serialise.hh index 5d9acf887..c29c6b29b 100644 --- a/src/libutil/serialise.hh +++ b/src/libutil/serialise.hh @@ -312,14 +312,14 @@ T readNum(Source & source) source(buf, sizeof(buf)); uint64_t n = - ((unsigned long long) buf[0]) | - ((unsigned long long) buf[1] << 8) | - ((unsigned long long) buf[2] << 16) | - ((unsigned long long) buf[3] << 24) | - ((unsigned long long) buf[4] << 32) | - ((unsigned long long) buf[5] << 40) | - ((unsigned long long) buf[6] << 48) | - ((unsigned long long) buf[7] << 56); + ((uint64_t) buf[0]) | + ((uint64_t) buf[1] << 8) | + ((uint64_t) buf[2] << 16) | + ((uint64_t) buf[3] << 24) | + ((uint64_t) buf[4] << 32) | + ((uint64_t) buf[5] << 40) | + ((uint64_t) buf[6] << 48) | + ((uint64_t) buf[7] << 56); if (n > std::numeric_limits<T>::max()) throw SerialisationError("serialised integer %d is too large for type '%s'", n, typeid(T).name()); diff --git a/src/libutil/split.hh b/src/libutil/split.hh new file mode 100644 index 000000000..d19d7d8ed --- /dev/null +++ b/src/libutil/split.hh @@ -0,0 +1,33 @@ +#pragma once + +#include <optional> +#include <string_view> + +#include "util.hh" + +namespace nix { + +// If `separator` is found, we return the portion of the string before the +// separator, and modify the string argument to contain only the part after the +// separator. Otherwise, wer return `std::nullopt`, and we leave the argument +// string alone. +static inline std::optional<std::string_view> splitPrefixTo(std::string_view & string, char separator) { + auto sepInstance = string.find(separator); + + if (sepInstance != std::string_view::npos) { + auto prefix = string.substr(0, sepInstance); + string.remove_prefix(sepInstance+1); + return prefix; + } + + return std::nullopt; +} + +static inline bool splitPrefix(std::string_view & string, std::string_view prefix) { + bool res = hasPrefix(string, prefix); + if (res) + string.remove_prefix(prefix.length()); + return res; +} + +} diff --git a/src/libutil/topo-sort.hh b/src/libutil/topo-sort.hh new file mode 100644 index 000000000..7a68ff169 --- /dev/null +++ b/src/libutil/topo-sort.hh @@ -0,0 +1,40 @@ +#include "error.hh" + +namespace nix { + +template<typename T> +std::vector<T> topoSort(std::set<T> items, + std::function<std::set<T>(const T &)> getChildren, + std::function<Error(const T &, const T &)> makeCycleError) +{ + std::vector<T> sorted; + std::set<T> visited, parents; + + std::function<void(const T & path, const T * parent)> dfsVisit; + + dfsVisit = [&](const T & path, const T * parent) { + if (parents.count(path)) throw makeCycleError(path, *parent); + + if (!visited.insert(path).second) return; + parents.insert(path); + + std::set<T> references = getChildren(path); + + for (auto & i : references) + /* Don't traverse into items that don't exist in our starting set. */ + if (i != path && items.count(i)) + dfsVisit(i, &path); + + sorted.push_back(path); + parents.erase(path); + }; + + for (auto & i : items) + dfsVisit(i, nullptr); + + std::reverse(sorted.begin(), sorted.end()); + + return sorted; +} + +} diff --git a/src/libutil/util.cc b/src/libutil/util.cc index a0a8ff4d3..c0b9698ee 100644 --- a/src/libutil/util.cc +++ b/src/libutil/util.cc @@ -374,7 +374,7 @@ void writeLine(int fd, string s) } -static void _deletePath(int parentfd, const Path & path, unsigned long long & bytesFreed) +static void _deletePath(int parentfd, const Path & path, uint64_t & bytesFreed) { checkInterrupt(); @@ -414,7 +414,7 @@ static void _deletePath(int parentfd, const Path & path, unsigned long long & by } } -static void _deletePath(const Path & path, unsigned long long & bytesFreed) +static void _deletePath(const Path & path, uint64_t & bytesFreed) { Path dir = dirOf(path); if (dir == "") @@ -435,12 +435,12 @@ static void _deletePath(const Path & path, unsigned long long & bytesFreed) void deletePath(const Path & path) { - unsigned long long dummy; + uint64_t dummy; deletePath(path, dummy); } -void deletePath(const Path & path, unsigned long long & bytesFreed) +void deletePath(const Path & path, uint64_t & bytesFreed) { //Activity act(*logger, lvlDebug, format("recursively deleting path '%1%'") % path); bytesFreed = 0; @@ -494,6 +494,7 @@ std::pair<AutoCloseFD, Path> createTempFile(const Path & prefix) { Path tmpl(getEnv("TMPDIR").value_or("/tmp") + "/" + prefix + ".XXXXXX"); // Strictly speaking, this is UB, but who cares... + // FIXME: use O_TMPFILE. AutoCloseFD fd(mkstemp((char *) tmpl.c_str())); if (!fd) throw SysError("creating temporary file '%s'", tmpl); @@ -1449,7 +1450,7 @@ string base64Decode(std::string_view s) char digit = decode[(unsigned char) c]; if (digit == -1) - throw Error("invalid character in Base64 string"); + throw Error("invalid character in Base64 string: '%c'", c); bits += 6; d = d << 6 | digit; diff --git a/src/libutil/util.hh b/src/libutil/util.hh index 630303a5d..3a20679a8 100644 --- a/src/libutil/util.hh +++ b/src/libutil/util.hh @@ -125,7 +125,7 @@ void writeLine(int fd, string s); second variant returns the number of bytes and blocks freed. */ void deletePath(const Path & path); -void deletePath(const Path & path, unsigned long long & bytesFreed); +void deletePath(const Path & path, uint64_t & bytesFreed); std::string getUserName(); |