Where in Go's source code can I find their implementation of make.
Turns out the "code search" functionality is almost useless for such a central feature of the language, and I have no good way to determine if I should be searching for a C function, a Go function, or what.
Also in the future how do I figure out this sort of thing without resorting to asking here? (i.e.: teach me to fish)
EDIT
P.S. I already found http://golang.org/pkg/builtin/#make, but that, unlike the rest of the go packages, doesn't include a link to the source, presumably because it's somewhere deep in compiler-land.
There is no make() as such. Simply put, this is happening:
go code: make(chan int)
symbol substitution: OMAKE
symbol typechecking: OMAKECHAN
code generation: runtime·makechan
gc, which is a go flavoured C parser, parses the make call according to context (for easier type checking).
This conversion is done in cmd/compile/internal/gc/typecheck.go.
After that, depending on what symbol there is (e.g., OMAKECHAN for make(chan ...)),
the appropriate runtime call is substituted in cmd/compile/internal/gc/walk.go. In case of OMAKECHAN this would be makechan64 or makechan.
Finally, when running the code, said substituted function in pkg/runtime is called.
How do you find this
I tend to find such things mostly by imagining in which stage of the process this
particular thing may happen. In case of make, with the knowledge that there's no
definition of make in pkg/runtime (the most basic package), it has to be on compiler level
and is likely to be substituted to something else.
You then have to search the various compiler stages (gc, *g, *l) and in time you'll find
the definitions.
As a matter of fact make is a combination of different functions, implemented in Go, in the runtime.
makeslice for e.g. make([]int, 10)
makemap for e.g. make(map[string]int)
makechan for e.g. make(chan int)
The same applies for the other built-ins like append and copy.
Related
I mostly use constants for documentation purposes e.g. a useful variable name or when I repeat certain sequences of strings over and over and don't want to change them manually. But I was wondering whether there's any performance difference. Am I right in my assumptions that there's no runtime difference between a literal and a constant, since constants are replaced at runtime?
Maybe I am misunderstanding, but I didn't find anything that tells me that this is wrong. The Go Tour doesn't provide any valuable information on and nor did the Constants blog post.
There's nothing that says one way or another whether even this trivial program:
package main
func main() {}
might run fast as lightning when compiled on a Tuesday, but slow as molasses when compiled on a late Friday afternoon. (Perhaps the Go compiler is anxious to head home for a beer and a weekend off and produced terrible code on Friday afternoon.1)
That said, if you're comparing, e.g.:
package main
import (
"fmt"
)
const hello = "hello"
var playground = "playground"
func main() {
fmt.Printf("%s, %s\n", hello, playground)
}
we might note that in the const variant (hello), the compiler is forced to know at compile time that the string literal "hello" is a string literal, while in the var variant (playground), the compiler could be lazy and assume that the variable playground might be modified in some other function. This in turn, combined with the ability of the compiler to know that fmt.Println is a particular function—the way GCC inserts special knowledge of the C printf function, for instance—could allow the compiler to more easily compile this to:
fmt.Printf("hello, %s\n", playground)
where only one runtime reflect happens, in case the variable playground has changed. But the existing Go compilers use SSA (see also https://golang.org/pkg/cmd/compile/internal/ssa/) and there are no writes to the variable, so we can expect simple (and usually simple = fast) runtime code here.
Playing with the Godbolt compiler site, it seems that when using const, the current compiler actually has to insert one conversion to string. The var version winds up with less runtime code. I didn't test it with string literals inserted. The %s directives are never expanded in line, but fmt.Printf really calls fmt.Fprintf directly, with os.Stdout as the first argument.
Overall, you're usually best off writing the clearest code you can. Then, if it's too slow (for whatever definition you have of "too slow"), measure. I'm guilty of overdoing my coding-time optimization myself, though. :-)
1Don't anthropomorphize computers. They hate that!
What is the most idiomatic way to get some form of context to the yacc parser in goyacc, i.e. emulate the %param command in traditional yacc?
I need to parse to my .Parse function some context (in this case including for instance where to build its parse tree).
The goyacc .Parse function is declared
func ($$rcvr *$$ParserImpl) Parse($$lex $$Lexer) int {
Things I've thought of:
$$ParserImpl cannot be changed by the .y file, so the obvious solution (to add fields to it) is right out, which is a pity.
As $$Lexer is an interface, I could stuff the parser context into the Lexer implementation, then force type convert $$lex to that implementation (assuming my parser always used the same lexer), but this seems pretty disgusting (for which read non-idiomatic). Moreover there is (seemingly) no way to put a user-generated line at the top of the Parse function like c := yylex.(*lexer).c, so in the many tens of places I want to refer to this variable, I have to use the rather ugly form yylex.(*lexer).c rather than just c.
Normally I'd use %param in normal yacc / C (well, bison anyway), but that doesn't exist in goyacc.
I'd like to avoid postprocessing my generated .go file with sed or perl for what are hopefully obvious reasons.
I want to be able to (go)yacc parse more than one file at once, so a global variable is not possible (and global variables are hardly idiomatic).
What's the most idiomatic solution here? I keep thinking I must be missing something simple.
My own solution is to modify goyacc (see this PR) which adds a %param directive allowing one or more fields to be added to the $$ParserImpl structure (accessible as $$rcvr in code). This seems the most idiomatic route. This permits not only passing context in, but the ability for the user to add additional func()s using $$ParserImpl as a receiver.
With the following code, go vet does not show an "out of bounds" error as I would expect:
package main
func main() {
a := make([]string, 1)
a[2] = "foo"
}
From the go vet documentation:
Flag: -shift
Shifts equal to or longer than the variable's length.
If go vet is not the tool to catch these errors, what is? Compiling and/or testing the code will catch this, but I'm looking for a static analysis based tool.
Its true that Go vet is for catching suspicious runtime error, by using some heuristics. The first Para is exact regarding its work
Vet examines Go source code and reports suspicious constructs, such as Printf calls whose arguments do not align with the format string. Vet uses heuristics that do not guarantee all reports are genuine problems, but it can find errors not caught by the compilers.
also in documentation its mentioned that
Note that the tool does not check every possible problem and depends on unreliable heuristics.
also the code which you are using to check for vetting your package is something very difficult to find by those heuristics as you are using a dynamic slice which can be appended or modified at runtime.
thereby not a perfect heuristic can be thought about for that.
fmt.Printf("%d", "scsa", "DSD")
those heuristic can catch things like this it all depends on what the training data is.
So vet should be a tool to take a quick look whether there is some general mistake which has been missed by you (If It gets caught :-) )its nothing like a compile tool or runtime checker it just runs some heuristics on the plane code you have written.
also documentation provides a list of available checks some examples are
Assembly declarations,
Copying locks,
Printf family,
Methods,
Struct tags,
etc there are many, you can see and read the complete documentation
I know this might be controversial or not very broad but I'm going to try to be very specific and relate to other questions.
Ok so when I make a Go program what things should I be thinking in terms of how I should organize my project? (E.g. should I think ok I'm going to have controllers of some sort so I should have a controller subdirectory that's going to do this so I should have that)
How should I structure a package?
For example the current program I'm working on I'm trying to make a SteamBot using this package
But while I'm writing it I don't know if I should export certain methods into their own own files, e.g. I'd have something like
func (t *tradeBot) acceptTrade() {}
func (t *tradeBot) declineTrade() {}
func (t *tradeBot) monitorTrade() {}
func (t *tradeBot) sendTrade() {}
each method is going to have quite a bit of code so should I export each method into its own file or just have 1 file with 3000 lines of code in it?
Also using global variables so that I can set one variable and then leave it and be able to use it in multiple functions, or is this bad and I should pass the variables as arguments?
Also should I order my files like:
package
imports
constants
variables
functions
methods
Or do I just put things where I need them?
The first place to look for inspiration is the Go standard library. The Go standard library is a pretty good guide of what "good" Go code should look like. A library isn't quite the same as an application, but it's definitely a good introduction.
In general, you would not break up every method into its own file. Go tends to prefer larger files that cover a topic. 3000 lines seems quite large, though. Even the net/http server.go is only 2200 lines, and it's pretty big.
Global mutable variables are just as bad in Go as in any language. Since Go relies so heavily on concurrent programming, global mutable variables are quite horrible. Don't do that. One common exception is sync structures like sync.Pool or sync.Once, which sometimes are package global, but are also designed to be accessed concurrently. There are also sometimes "Default" versions of structures, like http.DefaultClient, but you can still pass explicit ones to functions. Again, look through the Go standard library and see what is common and what is rare.
Just a few tips that you hopefully find useful:
Organize code into multiple files and packages by features, not by layers. This becomes more important the bigger your application becomes. package controllers with one or two controllers is probably ok, but putting hundreds of unrelated controllers in the same package doesn't make any sense. The following article explains it very well: http://www.javapractices.com/topic/TopicAction.do?Id=205
Global variables sometimes make code easier to write however they should be used with care. I think unexported global variables for things like logger, debug flags, etc are ok.
In all the projects I've worked with in other languages the bracing-style of choice has been the Allman Style(aka ANSI style). The lack of a free-form bracing style(parenthesis too) is something I miss from other C-style syntax family of languages when working in Go.
Can anyone come up with a way to make the Go compiler accept the following bracing-style?
package main
import "fmt"
func main()
{
f()
fmt.Println("Returned normally from f.")
}
func f()
{
fmt.Println("In function f.")
}
Note I am aware of the arguments for why Go was designed with such artificial 'limitation', but I'm not really buying into it. I'm a firm believer that the bracing-style used should really be decided by the coding-standard adopted by the people or company working on the code-base rather than being forced upon by the language itself.
As such please consider my question within the scope of 'how it can be done' rather than 'why not to do it and just adapt'.
Thanks
I double the braces up.
if x < 0 {
{
return sqrt(-x) + "i"
}}
It's not ideal but better than trying to scan columns 1-120 for matching braces.
This may not be exactly what you are looking for, but it is one possible way.
You could write a 'translator program,' essentially an incredibly simple compiler that converts from what you wrote, effectively a Go variant, to what the Go compiler itself expects.
You could do that with something along the lines of a program, even shell script, that applies the regular expression 's/(\n)$^\s*{/{\1/g' to the entire program (though it would need to look at the full string of the file and not break it up line-by-line, so you couldn't just pass that expression as an argument to sed, for example). Then write the converted file out to a temporary one, and run the Go compiler on it.
This method has the advantage of not requiring any changes to Go itself, though the disadvantage is that your files will not compile without your extra script. If you normally use a Makefile, this is probably fine, but sometimes could still be inconvenient.
Succinctly, no. The language is (or was a year or more ago) defined with semi-colons, and the compiler inserts semi-colons implicitly at the ends of lines - unless there's a brace at the end of the line. So, if you write a condition (which doesn't need the parentheses, please note) and do not put an open brace at the end of the line, then the Go compiler inserts one for you - leaving a null statement after the if, and the braces enclosing a block that is executed unconditionally.
#epw suggests a reformatter; I think that is a reasonable suggestion. Go comes with gofmt to convert to the canonical Go style. You'd have to write something to convert from canonical to Allman style, and vice versa, and ensure that you pre-compile your Go-Allman source into Go-canonical format before compiling it to object files. On the whole, this is more traumatic than accepting that Go has its own rules which are no more eccentric than Python's rules and that it is simplest to go with the flow and accept that coding in Go involves coding in non-Allman (approximately K&R) style. (I like and use Allman style in C; I use Go-style in Go.)
Give a try to https://gofork.org
forkgo supports allman/horstmann style:
package main
import
( "fmt"
)
func main()
{ if false
{ fmt.Println("jack")
fmt.Println("forkgo")
} else
{ fmt/
.Println("hello")
fmt.Println("forkgo")
}
}