diff options
author | Jo-Philipp Wich <jo@mein.io> | 2023-08-12 01:44:06 +0200 |
---|---|---|
committer | GitHub <noreply@github.com> | 2023-08-12 01:44:06 +0200 |
commit | 306da3c30ad9a334ba60e6e8431726b1220b8ccd (patch) | |
tree | 6839460fa242e64474ae98f2ccea2072fd1d0d3e /docs/LMO.md | |
parent | 4c3584bbdc1abe377c4b85099740f61d3213c1c5 (diff) | |
parent | 155d2bdd6ed787c553a1f338ee77478a1e3a1262 (diff) |
Merge pull request #6513 from stokito/luci-docs
docs: Synchronize with Wiki
Diffstat (limited to 'docs/LMO.md')
-rw-r--r-- | docs/LMO.md | 159 |
1 files changed, 82 insertions, 77 deletions
diff --git a/docs/LMO.md b/docs/LMO.md index 961a45ba84..3b7b5f92ae 100644 --- a/docs/LMO.md +++ b/docs/LMO.md @@ -1,7 +1,12 @@ -LMO is a simple binary format to pack language strings into a more efficient form. Although it's suitable to store any kind of key-value table, it's only used for the LuCI *.po based translation system at the moment. The abbreviation "LMO" stands for "Lua Machine Objects" in the style of the GNU gettext *.mo format. +# LMO - Lua Machine Objects +See [online wiki](https://github.com/openwrt/luci/wiki/LMO) for latest version. -# Format Specification +LMO is a simple binary format to pack language strings into a more efficient form. +Although it's suitable to store any kind of key-value table, it's only used for the LuCI *.po based translation system at the moment. +The abbreviation "LMO" stands for "Lua Machine Objects" in the style of the GNU gettext *.mo format. + +## Format Specification A LMO file is divided into two parts: the payload and the index lookup table. All segments of the file are 4 Byte aligned to ease reading and processing of the format. @@ -50,95 +55,95 @@ Schema: -# Processing +## Processing In order to process a LMO file, an implementation would have to do the following steps: -## Read Index +### Read Index 1. Locate and open the archive file -1. Seek to end of file - 4 bytes (sizeof(uint32_t)) -1. Read 32bit index offset and swap from network to native byte order -1. Seek to index offset, calculate index length: filesize - index offset - 4 -1. Initialize a linked list for index table entries -1. Read each index entry until the index length is reached, read and byteswap 4 * uint32_t for each step -1. Seek to begin of file +2. Seek to end of file - 4 bytes (sizeof(uint32_t)) +3. Read 32bit index offset and swap from network to native byte order +4. Seek to index offset, calculate index length: filesize - index offset - 4 +5. Initialize a linked list for index table entries +6. Read each index entry until the index length is reached, read and byteswap 4 * uint32_t for each step +7. Seek to begin of file -## Read Entry +### Read Entry 1. Calculate the unsigned 32bit hash of the entries key value (see "Hash Function" section below) -1. Obtain the archive index -1. Iterate through the linked index list, perform the following steps for each entry: - 1. Compare the entry hash value with the calculated hash from step 1 - 2. If the hash values are equal proceed with step 4 - 3. Select the next entry and repeat from step 3.1 -1. Seek to the file offset specified in the selected entry -1. Read as much bytes as specified in the entry length into a buffer -1. Return the buffer value +2. Obtain the archive index +3. Iterate through the linked index list, perform the following steps for each entry: + 1. Compare the entry hash value with the calculated hash from step 1 + 2. If the hash values are equal proceed with step 4 + 3. Select the next entry and repeat from step 3.1 +4. Seek to the file offset specified in the selected entry +5. Read as much bytes as specified in the entry length into a buffer +6. Return the buffer value -# Hash Function +## Hash Function The current LuCI-LMO implementation uses the "Super Fast Hash" function which was kindly put in the public domain by it's original author. See http://www.azillionmonkeys.com/qed/hash.html for details. Below is the C-Implementation of this function: - - #if (defined(__GNUC__) && defined(__i386__)) - #define sfh_get16(d) (*((const uint16_t *) (d))) - #else - #define sfh_get16(d) ((((uint32_t)(((const uint8_t *)(d))[1])) << 8)\ - +(uint32_t)(((const uint8_t *)(d))[0]) ) - #endif - - uint32_t sfh_hash(const char * data, int len) - { - uint32_t hash = len, tmp; - int rem; - - if (len <= NULL) return 0; - - rem = len & 3; - len >>= 2; - - /* Main loop */ - for (;len > 0; len--) { - hash += sfh_get16(data); - tmp = (sfh_get16(data+2) << 11) ^ hash; - hash = (hash << 16) ^ tmp; - data += 2*sizeof(uint16_t); - hash += hash >> 11; - } - - /* Handle end cases */ - switch (rem) { - case 3: hash += sfh_get16(data); - hash ^= hash << 16; - hash ^= data[sizeof(uint16_t)] << 18; - hash += hash >> 11; - break; - case 2: hash += sfh_get16(data); - hash ^= hash << 11; - hash += hash >> 17; - break; - case 1: hash += *data; - hash ^= hash << 10; - hash += hash >> 1; - } - - /* Force "avalanching" of final 127 bits */ - hash ^= hash << 3; - hash += hash >> 5; - hash ^= hash << 4; - hash += hash >> 17; - hash ^= hash << 25; - hash += hash >> 6; - - return hash; +```c +#if (defined(__GNUC__) && defined(__i386__)) +#define sfh_get16(d) (*((const uint16_t *) (d))) +#else +#define sfh_get16(d) ((((uint32_t)(((const uint8_t *)(d))[1])) << 8)\ + +(uint32_t)(((const uint8_t *)(d))[0]) ) +#endif + +uint32_t sfh_hash(const char * data, int len) +{ + uint32_t hash = len, tmp; + int rem; + + if (len <= 0 || data == NULL) return 0; + + rem = len & 3; + len >>= 2; + + /* Main loop */ + for (;len > 0; len--) { + hash += sfh_get16(data); + tmp = (sfh_get16(data+2) << 11) ^ hash; + hash = (hash << 16) ^ tmp; + data += 2*sizeof(uint16_t); + hash += hash >> 11; } - -# Reference Implementation + /* Handle end cases */ + switch (rem) { + case 3: hash += sfh_get16(data); + hash ^= hash << 16; + hash ^= data[sizeof(uint16_t)] << 18; + hash += hash >> 11; + break; + case 2: hash += sfh_get16(data); + hash ^= hash << 11; + hash += hash >> 17; + break; + case 1: hash += *data; + hash ^= hash << 10; + hash += hash >> 1; + } + + /* Force "avalanching" of final 127 bits */ + hash ^= hash << 3; + hash += hash >> 5; + hash ^= hash << 4; + hash += hash >> 17; + hash ^= hash << 25; + hash += hash >> 6; + + return hash; +} +``` + +## Reference Implementation A reference implementation can be found here: -http://luci.subsignal.org/trac/browser/luci/trunk/libs/lmo/src +https://github.com/openwrt/luci/blob/master/modules/luci-base/src/template_lmo.c -The lmo_po2lmo.c executable implements a *.po to *.lmo conversation utility and lmo_lookup.c is a simple *.lmo test utility. -Lua bindings for lmo are defined in lmo_lualib.c and associated headers. +The `po2lmo.c` executable implements a `*.po` to `*.lmo` conversation utility. +Lua bindings for lmo are defined in `template_lualib.c` and associated headers. |