ff063d28c6
We do this, because the changes in this patch fundamentally alter the way code is generated by the tool. This will, in some cases, be incompatible with older versions. * Performs cleanup and minor code fixes. * Adds two new command line flags: * -prefix: This accepts a partial path. It is used when generating a target function name, as well as the key for the Table of Contents map (see below). The specified prefix is applied to the input file name, causing the prefix section to be stripped from the input file path. E.g.: ``` input: /path/to/foo.x prefix: /path/to output: /foo.x ``` * -toc: This is a boolean flag which tells the tool to generate a table of contents for the generated data files. It creates a separate 'bindata-toc.go' file, which defines a global map named `go_bindata`. It then appends an `init` function to the generated file. This function makes the data function register itself with the global map. * Fixes the function names the tool infers from input file names to include the full path. This fixes potential name collisions when the same file name is processed from different directories. For example, we can now safely import the following two files: ``` input file: css/ie/foo.css output function: css_ie_foo_css() input file: css/chrome/foo.css output function: css_chrome_foo_css() ``` |
||
---|---|---|
testdata | ||
bytewriter.go | ||
CONTRIBUTORS | ||
LICENSE | ||
main.go | ||
README.md | ||
stringwriter.go | ||
toc.go | ||
translate.go | ||
version.go |
bindata
This tool converts any file into managable Go source code. Useful for embedding binary data into a go program. The file data is optionally gzip compressed before being converted to a raw byte slice.
Usage
The simplest invocation is to pass it only the input file name. The output file and code settings are inferred from this automatically.
$ go-bindata testdata/gophercolor.png
[w] No output file specified. Using 'testdata/gophercolor.png.go'.
[w] No package name specified. Using 'main'.
[w] No function name specified. Using 'testdata_gophercolor_png'.
This creates the testdata/gophercolor.png.go
file which has a package
declaration with name main
and one function named testdata_gophercolor_png
with
the following signature:
func testdata_gophercolor_png() []byte
You can now simply include the new .go file in your program and call
testdata_gophercolor_png()
to get the (uncompressed) image data. The function panics
if something went wrong during decompression. See the testdata directory for
example input and output files for various modes.
Aternatively, you can pipe the input file data into stdin. go-bindata
will
then spit out the generated Go code to stdout. This does require explicitly
naming the desired function name, as it can not be inferred from the
input data. The package name will still default to 'main'.
$ cat testdata/gophercolor.png | go-bindata -f gophercolor_png | gofmt
Invoke the program with the -h
flag for more options.
In order to strip off a part of the generated function name, we can use the -prefix
flag.
In the above example, the input file testdata/gophercolor.png
yields a function named
testdata_gophercolor_png
. If we want the testdata
component to be left out, we invoke
the program as follows:
$ go-bindata -prefix "testdata/" testdata/gophercolor.png
Lower memory footprint
Using the -nomemcopy
flag, will alter the way the output file is generated.
It will employ a hack that allows us to read the file data directly from
the compiled program's .rodata
section. This ensures that when we call
call our generated function, we omit unnecessary memcopies.
The downside of this, is that it requires dependencies on the reflect
and
unsafe
packages. These may be restricted on platforms like AppEngine and
thus prevent you from using this mode.
Another disadvantage is that the byte slice we create, is strictly read-only. For most use-cases this is not a problem, but if you ever try to alter the returned byte slice, a runtime panic is thrown. Use this mode only on target platforms where memory constraints are an issue.
The default behaviour is to use the old code generation method. This prevents the two previously mentioned issues, but will employ at least one extra memcopy and thus increase memory requirements.
For instance, consider the following two examples:
This would be the default mode, using an extra memcopy but gives a safe
implementation without dependencies on reflect
and unsafe
:
func myfile() []byte {
return []byte{0x89, 0x50, 0x4e, 0x47, 0x0d, 0x0a, 0x1a}
}
Here is the same functionality, but uses the .rodata
hack.
The byte slice returned from this example can not be written to without
generating a runtime error.
var _myfile = "\x89\x50\x4e\x47\x0d\x0a\x1a"
func myfile() []byte {
var empty [0]byte
sx := (*reflect.StringHeader)(unsafe.Pointer(&_myfile))
b := empty[:]
bx := (*reflect.SliceHeader)(unsafe.Pointer(&b))
bx.Data = sx.Data
bx.Len = len(_myfile)
bx.Cap = bx.Len
return b
}
Optional compression
When the -uncompressed
flag is given, the supplied resource is not GZIP compressed
before being turned into Go code. The data should still be accessed through
a function call, so nothing changes in the usage of the generated file.
This feature is useful if you do not care for compression, or the supplied resource is already compressed. Doing it again would not add any value and may even increase the size of the data.
The default behaviour of the program is to use compression.
Table of Contents
With the -toc
flag, we can have go-bindata
create a table of contents for all the files
which have been generated by the tool. It does this by first generating a new file named
bindata-toc.go
. This contains a global map of type map[string] func() []byte
. It uses the
input filename as the key and the data function as the value. We can use this
to fetch all data for our files, matching a given pattern.
It then appands an init
function to each generated file, which simply makes the data
function append itself to the global bindata
map.
Once you have compiled your program with all these new files and run it, the map will be populated by all generated data files.
Note: The bindata-toc.go
file will not be created when we run in pipe
mode.
The reason being, that the tool does not write any files at all, as it has no idea
where to save them. The data file is written to stdout
instead after all.
Table of Contents keys
The keys used in the go_bindata
map, are the same as the input file name passed to go-bindata
.
This includes the fully qualified (absolute) path. In most cases, this is not desireable, as it
puts potentially sensitive information in your code base. For this purpose, the tool supplies
another command line flag -prefix
. This accepts a portion of a path name, which should be
stripped off from the map keys and function names.
For example, running without the -prefix
flag, we get:
$ go-bindata /path/to/templates/foo.html
go_bindata["/path/to/templates/foo.html"] = path_to_templates_foo_html
Running with the -prefix
flag, we get:
$ go-bindata -prefix "/path/to/" /path/to/templates/foo.html
go_bindata["templates/foo.html"] = templates_foo_html
bindata-toc.go
The bindata-toc.go
file is very simple and looks as follows:
package $PACKAGENAME
// Global Table of Contents map. Generated by go-bindata.
// After startup of the program, all generated data files will
// put themselves in this map. The key is the full filename, as
// supplied to go-bindata.
var go_bindata = make(map[string] func() []byte)