Partial draft of Programming in Go for initial review #141

ksbhaskar · 2018-06-24T19:13:45Z

Utility function wrapper descriptions yet to be added. This is a request for an initial review, not a review before merging.

communikit · 2018-06-25T13:05:35Z

MultiLangProgGuide/MultiLangProgGuide.rst

+<https://golang.org/cmd/cgo/>`_ to provide the Go programmer with a
+YottaDB package. YottaDB C functions are wrapped by Go methods where a
+method can meaningfully be associated with a Go structure, and Go
+functions otherwise.


".... and otherwise associated with a Go function"?

Changed to “and by Go functions otherwise” to tie the phrase to “wrapped”.

communikit · 2018-06-25T13:06:10Z

MultiLangProgGuide/MultiLangProgGuide.rst

+method can meaningfully be associated with a Go structure, and Go
+functions otherwise.
+
+As the Go language has important differences from C (for example it


Comma after "for example"

communikit · 2018-06-25T13:08:59Z

MultiLangProgGuide/MultiLangProgGuide.rst

+lacks macros, but has methods), below are Go-specific sections of the
+`Quick Start`_, `Concepts`_, `Symbolic Constants`_, `Data Structures &
+Type Definitions`_ and `Simple API`_ sections above. The sections
+below specific to Go are intended to supplement their general (C)


"The sections specific to Go below" or "The sections below that are specific to Go"

shabiel · 2018-06-25T13:13:43Z

MultiLangProgGuide/MultiLangProgGuide.rst

+3. Put your GO program in the :code:`$ydb_dir` directory, XYZ
+   (instructions to include headers).  As a sample program,
+   you can download the `wordfreq.go
+   <https://raw.githubusercontent.com/YottaDB/YottaDBtest/master/simpleapi/inref/wordfreq.c>`_


This link points to the C code.

True. The Go code doesn't exist yet because the Go wrapper doesn't exist yet. We're drafting the user documentation to specify the API.

estess · 2018-06-25T14:32:56Z

MultiLangProgGuide/MultiLangProgGuide.rst

+  - Empty string subscripts precede all numeric subscripts. By
+    default, YottaDB prohibits empty string subscripts for global
+    variables but permits them for local variables (see `Local and
+    Global Variables`_). *Note: YottaDB recommends against


Perhaps should recommend against the practice of using null vars in applications instead of recommending against applications?

estess · 2018-06-25T14:36:56Z

MultiLangProgGuide/MultiLangProgGuide.rst

+Programming YottaDB in the `Go language <https://golang.org/>`_ is
+accomplished through a wrapper that uses `cgo
+<https://golang.org/cmd/cgo/>`_ to provide the Go programmer with a
+YottaDB package. YottaDB C functions are wrapped by Go methods where a


If this is the name of the actual package, the package name should be all lowercase. Otherwise it should be clear that it is a package 'to support the use of YottaDB'.

Rephrased as: to provide a “yottadb” package for access to YottaDB from Go application code.

estess · 2018-06-25T14:41:10Z

MultiLangProgGuide/MultiLangProgGuide.rst

+below specific to Go are intended to supplement their general (C)
+counterparts, rather than stand-alone sections.
+
+Application programs should not directly use the YottaDB C API


This cannot be overstated enough. Direct calls to the YottaDB C routines (both SimpleAPI and especially the utility routines), are NOT supported as they bypass some very important controls.

Rephrased as: Application programs must not directly use the YottaDB C API structures and functions (those prefixed by :code:C.) as such usage bypasses important controls, but should instead use the structures, methods and functions exposed by the YottaDB Go wrapper

estess · 2018-06-25T14:42:43Z

MultiLangProgGuide/MultiLangProgGuide.rst

+(Installing YottaDB), install the Go wrapper:
+
+Download the Go wrapper from XYZ [provide URL when released].  Unpack
+the contents in its own directory (e.g, $HOME/go/src/yottadb, and


Missing close paren before comma

estess · 2018-06-25T15:04:14Z

MultiLangProgGuide/MultiLangProgGuide.rst

+Go Concepts
+===========
+
+XYZ - Add anything special that a Go programmer should know before


Likely need to add something here about changing yottadb.go since it has direct references to where YDB is installed. These cannot currently be environment variables so stuck with fixing up the cgo comment lines that define the macro include and link parameters (-I, -l -L). The cgo information says use of a tool called "pkg-config" (available in ubuntu repository) can get around the hard-coded filerefs where things are not present in the build directory. I have not used pkg-config so have no idea how to use it but it would keep us from having to suggest source modifications before building.

OK, will add when you provide suitable draft text.

estess · 2018-06-25T16:37:33Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+.. code-block:: go
+
+    DeleteS(deltype int) int


In order to allow use of the C.YDB_DEL_NODE/DEL_TREE constants to be used in this method, the type of 'deltype' should probably be C.int.

Looks like this change was undone which is good since we discovered C #define values can be treated as ints instead of requiring them to be C.int.

estess · 2018-06-25T16:39:00Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+    IncrS(incr, retval *BufferT) int
+
+The :code:`IncrS()` method wraps `ydb_incr_s()`_ to atomically


If incr is either unallocated or is nil, it takes on the default value of 1.

Added: If :code:incr is unallocated or the :code:len_usedfield is zero, the default increment is 1.

estess · 2018-06-25T16:43:32Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+.. code-block:: go
+
+    LockS( timeoutNsec uint64, ... *KeyT) int


Space before timeoutNsec. Also should be parm name prior to '...' like "locklist" or some such.

Might want to note maximum here of ((#locks * 3) + 1) must be <= 36.

Fixed. Also noted that the number of locks ≤11 (easier to parse than the formula).

estess · 2018-06-25T16:47:40Z

MultiLangProgGuide/MultiLangProgGuide.rst

+====================
+
+The Go Comprehensive API is a project for the future, to follow the C
+`Comprehensive API`_


"follow the C Comprehensive API" -> "follow the C SimpleAPI" ?

A Go Comprehensive API would have to follow a C Comprehensive API.

estess · 2018-06-25T16:49:28Z

MultiLangProgGuide/MultiLangProgGuide.rst

+-----------------------
+
+GetSubBufAlloc(n)
+GetSubBufUsed(n)


Not understanding why these are "Utility Methods" instead of structure methods.

You read past where I had material to review. This is an area where I made notes to myself. The next round of reviews will have something here worth reviewing.

ksbhaskar

Responses to collected comments by Steve Estes.

ksbhaskar · 2018-06-25T17:14:38Z

MultiLangProgGuide/MultiLangProgGuide.rst

+  - Empty string subscripts precede all numeric subscripts. By
+    default, YottaDB prohibits empty string subscripts for global
+    variables but permits them for local variables (see `Local and
+    Global Variables`_). *Note: YottaDB recommends against


ksbhaskar · 2018-06-25T17:18:49Z

MultiLangProgGuide/MultiLangProgGuide.rst

+Programming YottaDB in the `Go language <https://golang.org/>`_ is
+accomplished through a wrapper that uses `cgo
+<https://golang.org/cmd/cgo/>`_ to provide the Go programmer with a
+YottaDB package. YottaDB C functions are wrapped by Go methods where a


Rephrased as: to provide a “yottadb” package for access to YottaDB from Go application code.

ksbhaskar · 2018-06-25T17:40:16Z

MultiLangProgGuide/MultiLangProgGuide.rst

+below specific to Go are intended to supplement their general (C)
+counterparts, rather than stand-alone sections.
+
+Application programs should not directly use the YottaDB C API


Rephrased as: Application programs must not directly use the YottaDB C API structures and functions (those prefixed by :code:C.) as such usage bypasses important controls, but should instead use the structures, methods and functions exposed by the YottaDB Go wrapper

ksbhaskar · 2018-06-25T17:40:44Z

MultiLangProgGuide/MultiLangProgGuide.rst

+(Installing YottaDB), install the Go wrapper:
+
+Download the Go wrapper from XYZ [provide URL when released].  Unpack
+the contents in its own directory (e.g, $HOME/go/src/yottadb, and


ksbhaskar · 2018-06-25T17:41:53Z

MultiLangProgGuide/MultiLangProgGuide.rst

+Go Concepts
+===========
+
+XYZ - Add anything special that a Go programmer should know before


OK, will add when you provide suitable draft text.

ksbhaskar · 2018-06-25T19:59:24Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+.. code-block:: go
+
+    DeleteS(deltype int) int


ksbhaskar · 2018-06-25T20:02:35Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+    IncrS(incr, retval *BufferT) int
+
+The :code:`IncrS()` method wraps `ydb_incr_s()`_ to atomically


Added: If :code:incr is unallocated or the :code:len_usedfield is zero, the default increment is 1.

ksbhaskar · 2018-06-25T20:06:15Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+.. code-block:: go
+
+    LockS( timeoutNsec uint64, ... *KeyT) int


Fixed. Also noted that the number of locks ≤11 (easier to parse than the formula).

ksbhaskar · 2018-06-25T20:08:20Z

MultiLangProgGuide/MultiLangProgGuide.rst

+====================
+
+The Go Comprehensive API is a project for the future, to follow the C
+`Comprehensive API`_


A Go Comprehensive API would have to follow a C Comprehensive API.

ksbhaskar · 2018-06-25T20:09:25Z

MultiLangProgGuide/MultiLangProgGuide.rst

+-----------------------
+
+GetSubBufAlloc(n)
+GetSubBufUsed(n)


You read past where I had material to review. This is an area where I made notes to myself. The next round of reviews will have something here worth reviewing.

estess · 2018-06-27T14:08:03Z

MultiLangProgGuide/MultiLangProgGuide.rst

+Methods for each structure are classified as either `Go Access
+Methods`_ or `Go Simple API`_ methods. `Go Access Methods`_ are
+methods implemented in the Go wrapper for managing the structures for
+data interchange or `Go Access Methods`_ that wrap functionality


Is this second mention of "Go Access Methods" supposed to be "Go SimpleAPI Methods" ? Also looks like "methods" when speaking of "Go Simple API methods" is not part of the hilighted/quoted string like Methods is with "Go Access Methods" which looks odd.

Yes, the second one should be Go Simple API methods. Also, “methods” is not part of the hyperlink because the section header is “Go Simple API” which has methods and functions. Yes, it's a trifle inconsistent/awkward…

estess · 2018-06-27T14:08:08Z

MultiLangProgGuide/MultiLangProgGuide.rst

+  referencing the buffer.
+
+Set :code:`cbuft` in the :code:`BufferT`
+structure to reference the :code:`C.ydb_buffer_t` structure.


reference -> anchor?

Reference seems to be right, because cbuft is in Go memory but stores a pointer to a ydb_buffer_t structure in the YottaDB heap. So going from cbuft to the ydb_buffer_t would involve dereferencing a pointer.

estess · 2018-06-27T14:27:21Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+Set:
+
+- In each :code:`C.ydb_buffer_t` structure:


Should this reference be to a BufferT structure since the next one references the BufferTArray struct?

len_alloc and len_used are fields in a C.ydb_buffer_t structure, whereas cbuftary, elemsAlloc, and elemsUsed are fields in a BufferTArray structure. So the existing language seems correct to me.

estess · 2018-06-27T19:18:42Z

MultiLangProgGuide/MultiLangProgGuide.rst

+  :code:`C.YDB_ERR_INVZWRITECHAR`.
+- Otherwise, set the buffer referenced by :code:`buf_addr` to the
+  unencoded string, set :code:`len_used` to the length, and return
+  with a return code of :code:`C.YDB_OK`.

 Note that the length of a string in `zwrite format`_ is always less


Is this backwards? The original string is less than or equal to the zwrite format. The zwrite format can be much larger than the unencoded format.

Yes it is. Fixed.

estess · 2018-06-27T19:21:16Z

MultiLangProgGuide/MultiLangProgGuide.rst

+routine, in turn be passed to the Go function called by the glue
+routine. As :code:`tpfnparm` is passed from Go to YottaDB and back to
+Go, the memory it references should be allocated using
+`yottadb.malloc()`_ to protect it from the Go garbage collector.


"malloc" needs to be capitalized.

I have since changed the reference to the section header, which is "Go Malloc()".

estess · 2018-06-27T19:29:21Z

MultiLangProgGuide/MultiLangProgGuide.rst

-
-If there is an error, :code:`SubNextS()` has an `error return code`_
-prefixed with :code:`C.`.
+:code:`SubPrevtS()` wraps `ydb_subscript_previous_s()`_ to facilitate


SubPrevtS -> SubPrevS

estess · 2018-06-27T19:31:38Z

MultiLangProgGuide/MultiLangProgGuide.rst


-The function wraps `ydb_lock_s()`. It releases all `Locks`_ held by
+The function wraps `ydb_lock_s()`_. It releases all `Locks`_ held by
 the process, and acquires the locks whose resource names are specified


Perhaps this should spawn a Programming Consideration note - the effect of LockS() releasing existing locks is process wide, not thread wide or goroutine wide but process wide.

estess · 2018-06-27T19:33:32Z

MultiLangProgGuide/MultiLangProgGuide.rst

-----------------------
-Go KeyT Utility Methods
-----------------------
+------------


I thought we were getting rid of all the YDB prefixes on these routines? This was because the routine name is actually for example yottadb.YDBExit().

We did. You were commenting on a line that was there in the previous version and removed in the commit!

estess · 2018-06-27T19:37:45Z

MultiLangProgGuide/MultiLangProgGuide.rst

+Refer to the description of `ydb_stdout_stderr_adjust()`_, which this
+function wraps. The return code is :code:`C.YDB_OK`.
+
+Applications that do not mix M and non-M code do not need this


Is this true? I thought it was if stdout and stderr were being used by various components that one might want to use this routine. Use of this routine from Golang is currently an unknown. We don't know how Golang sets up its IO stuff. Would be interesting to see if using this in Golang then forking off a process caused any issues.

At least in the C world, it is true that applications which do not mix M and non-M code do not need this function. Only if simpleAPI code does a call-in is this function needed in C. As for GoLang, I expect it to be similar. If at all, this is needed, it would be when call-ins are done from GoLang. Are we planning on supporting that?

estess · 2018-06-27T19:38:54Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+The function wraps `ydb_timer_start()`_ to start a timer. Unless
+canceled, after the timer expires, YottaDB invokes the handler
+function referenced by 


Incomplete sentence.

That's because you were reviewing functions before I finished drafting the functions content. I had said to stop after Methods!

estess · 2018-06-27T19:47:49Z

MultiLangProgGuide/MultiLangProgGuide.rst

+it wraps. The YottaDB `Go ForkExec()`_ ensures that the child process
+is safely disconnected from YottaDB interprocess communication
+resources and database files.
+


I'm not at all sure that allowing users to specify ProcAttr is a good idea as our version of ForkExec() has a need to set it for YottaDB purposes too which may conflict with the the user. We would end up re-copying the array to add our element to it at best. Also don't like the ability it gives the user by bypass closing all the db files.

I don't think it will be acceptable to users to disable it, because ProcAttr includes SysProcAttr some of which are likely to be important to users. Besides, programs can always explicitly set environment variables using sys.Setenv(). So, I think we should just reserve environment variables with a “ydb_” prefix for our use, and document it (if it is not already).

That said, I wonder whether we should wrap syscall.StartProcess(), even though it just wraps syscall.ForkExec() since syscall.StartProcess() may be the os independent way of starting processes.

estess · 2018-06-27T19:55:24Z

MultiLangProgGuide/MultiLangProgGuide.rst

+its logic in a single goroutine.
+
+In order to avoid restricting Go applications from calling the YottaDB
+API from a single gorutine (which would be unnatural to a Go


gorutine -> goroutine

estess · 2018-06-27T19:56:07Z

MultiLangProgGuide/MultiLangProgGuide.rst

+API from a single gorutine (which would be unnatural to a Go
+programmer), the YottaDB Go wrapper includes logic that coerces the
+execution of the YottaDB runtime system to a single goroutine. Calling
+YottaDB C API functions bypasses this protection, and may result in


"directly" bypasses?

nars1 · 2018-06-28T19:59:41Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+    Free()
+
+The inverse of the :code:`Alloc()` method: release the bufer in


bufer -> buffer

nars1 · 2018-06-28T20:02:46Z

MultiLangProgGuide/MultiLangProgGuide.rst

+return value (zero if the structure has not yet been allocated, i.e.,
+:code:`cbuft` is :code:`nil`).
+
+If the :code:`len_used` field field of the :code:`C.ydb_buffer_t`


field field -> field

nars1 · 2018-06-28T20:04:59Z

MultiLangProgGuide/MultiLangProgGuide.rst

+:code:`cbuft` is :code:`nil`).
+
+If the :code:`len_used` field field of the :code:`C.ydb_buffer_t`
+structure is greater than the :code:`len_alloc` field (owing to a


owing to a -> for example, owing to a

Otherwise, the sentence makes it sound like even though len_used is greater than len_alloc, as long as a C.YDB_ERR_INVSTRLEN error was not issued, the return value from GetLenUsed() will not be C.YDB_ERR_INVSTRLEN.

I notice the "owing to a" usage is in various places so they might also need the above change. That said, I am fine if you decide not to incorporate this comment.

As discussed, these are left as is.

nars1 · 2018-06-28T20:10:26Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+.. code-block:: go
+
+    GetValBAry() *[]byte, int


Does the *[]byte syntax imply a Go slice (as opposed to an array)? If so, the Ary suffix in the function name might not be accurate. If not, I need to get up to speed on Go so never mind.

Also BAry stands for a Byte Array right?

Yes, BAry is Byte Array. @estess, please respond to the slice vs. array question. If it is a slice, the method name should be *BSlice presumably.

This is the address of a byte array. The wrapper will create the byte array from the C buffer and return the byte array's address.

nars1 · 2018-06-28T20:14:43Z

MultiLangProgGuide/MultiLangProgGuide.rst

+If the :code:`len_used` field of the :code:`C.ydb_buffer_t` structure
+referenced by :code:`cbuft` is greater than its :code:`len_alloc`
+field (owing to a prior :code:`C.YDB_ERR_INVSTRLEN` error), return
+:code:`len_alloc` bytes of the buffer referenced by the


Too many "referenced by" usages in this and the next line. Can we avoid one.

We discussed this, and describing a chain of pointers in English is what leads to the multiple usages. Improved language doesn't suggest itself to me at present.

nars1 · 2018-06-28T21:02:58Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+The inverse of the :code:`Alloc()` method: release the :code:`numSubs`
+buffers and the :code:`C.ydb_buffer_t` array. Set :code:`cbuftary` to
+:code:`nil`.


Will elemsAlloc and elemsUsed also be set to 0?

Technically not necessary, because if cbuftary is nil, elemsAlloc and elemsUsed are not meaningful. But I added zeroing them to the specification since that is good defensive programming.

estess · 2018-06-29T00:48:51Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+.. code-block:: go
+
+    SetVarStr(val *string) int


What are we calling it? SetValStr() or SetVarStr() ?? I think all the others are SetValStr() though thinking about it I have to wonder if Var() wouldn't have been a better choice.

It seems to me that for BufferT and BufferTArray, val is preferable to var because we are setting and getting values from a buffer. var would be preferable for KeyT methods, except that we leave it blank there.

estess · 2018-06-29T00:52:37Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+.. code-block:: go
+
+    SetUsed(newidx uint) int


Thinking on it, the newidx parm is misnamed. We aren't storing an index here. It is a count of active elements in the array. So perhaps 'newcnt' is a better name.

Changed to newUsed

estess · 2018-06-29T00:53:37Z

MultiLangProgGuide/MultiLangProgGuide.rst

+Use this method to set the current number of valid strings (subscripts
+or variable names) in the :code:`BufferTArray`.
+
+- If :code:`idx` is greater than :code:`elemsAlloc`, make no


This should reference 'newidx' (now 'newcnt') instead of 'idx'.

Changed from newidx to newUsed – should make it even clearer.

estess · 2018-06-29T00:54:11Z

MultiLangProgGuide/MultiLangProgGuide.rst

+- Otherwise, set :code:`elemsUsed` to :code:`newidx` and return with a
+  return code of :code:`C.YDB_OK`.
+
+Note that even if :code:`idx` is not greater than the value of


Another idx to change along with the newidx references.

Corrected to newUsed

estess · 2018-06-29T00:57:35Z

MultiLangProgGuide/MultiLangProgGuide.rst

+
+.. code-block:: go
+
+    DeleteS(deltype int) int


Looks like this change was undone which is good since we discovered C #define values can be treated as ints instead of requiring them to be C.int.

estess · 2018-06-29T01:08:34Z

MultiLangProgGuide/MultiLangProgGuide.rst

+The function wraps `ydb_init()`_ to initialize the YottaDB runtime
+system. This call is normally not required as YottaDB initializes
+itself on its first call, the exception being when an application
+wishes to set its own signal handlers (see


I haven't read the Signals chapter in awhile (on my list) but I'm not entire sure the need to call ydb_init() during signal setup is true anymore. The plan is to modify signal setup in YottaDB so that if a handler is defined, it is recorded. Then if that signal occurs, once we are done with YottaDB's processing of the signal, we will call the handler that was previously setup (Golang's handler). If handlers are replaced after YottaDB is initialized, yottadb won't know about it (except for SIGALRM which it specifically checks). I don't know if Golang plays well with others in terms of signals so best would be to have all the handlers setup before YottaDB is initialized. Note this handling is not Golang specific and would apply to an non-M main routine.

@nars1, I'll go with whatever you and @estess agree is right for the document.

estess · 2018-06-29T01:11:13Z

MultiLangProgGuide/MultiLangProgGuide.rst

+    // #include <errno.h>
+    // #include <string.h>
+    // #include "libyottadb.h"
+    // #cgo LDFLAGS: -L/usr/lib/yottadb/r122 -lyottadb -Wl,-rpath,/usr/lib/yottadb/r122


Probably should remove all the includes in this group except libyottadb.h as they have no relevance to what needs to be added. That was just what I had in my main routine for the POC. I removed them all and it still built.

Finished draft of Simple API methods and LockS() function. Need to draft description of wrapped utility functions. Edits in response to e-mail comments from Steve Estes June 21, 2018, 5:43pm EDT. Incorporated comments from Ranjani & Steve E. Reworked Programming in Go section through Methods for clarity & consistency. Yet to do: rework functions and draft programming considerations. Completed first draft of Programming in Go section. I believe I have updated the document with responses to all comments to date. With the possible exception of potential changes to yottadb.ForkExec(), the draft is complete and ready for another review pass. Restore inadvertently deleted .gitignore file and fix typo in MLPG. Fix .gitignore Fix a typo, and add KeyT access methods that invoke the corresponding access methods on members of the KeyT structure. Fix typo Includes responses to Narayanan's reviews through this time. Added yottadb.Release(), KeyT methods. Removed StdoutStdErrAdjust(). Responded to comments and fixed typos. Edits to remove typos and improve readability.

…d errors.

…tion of interface is a to do.

…r interface has a string return.

…l, and related cleanup of language.

ksbhaskar added the enhancement label Jun 24, 2018

ksbhaskar requested review from communikit and estess June 24, 2018 19:13

communikit reviewed Jun 25, 2018

View reviewed changes

shabiel reviewed Jun 25, 2018

View reviewed changes

estess suggested changes Jun 25, 2018

View reviewed changes

ksbhaskar commented Jun 25, 2018

View reviewed changes

estess suggested changes Jun 27, 2018

View reviewed changes

nars1 requested changes Jun 28, 2018

View reviewed changes

estess suggested changes Jun 29, 2018

View reviewed changes

ksbhaskar force-pushed the working branch from ade7cf0 to 5cc7417 Compare June 29, 2018 21:38

ksbhaskar added 9 commits June 29, 2018 18:16

Fix formatting issues from merge with current MLPG

cb63057

Fix a few last minute typos

7d9bc6d

Added BufferT methods SetLenUsed() and SetValStrLit(). Fixed typos an…

48fc7f9

…d errors.

Added BufferTArray method SetLenUsed()

7daf1f9

Added method BufferTArray SetValStrLit()

4fe4f01

Change return codes to use an error interface. Noted detailed descrip…

29214c9

…tion of interface is a to do.

Provide some details of error interface.

5f12578

Futhre changes to error handling to reflect the fact that the Go erro…

de37213

…r interface has a string return.

Change Go error return codes to just error return; empty string to ni…

f62b6a0

…l, and related cleanup of language.


		IncrS(incr, retval *BufferT) int

		The :code:`IncrS()` method wraps `ydb_incr_s()`_ to atomically


		Free()

		The inverse of the :code:`Alloc()` method: release the bufer in

Partial draft of Programming in Go for initial review #141

Are you sure you want to change the base?

Partial draft of Programming in Go for initial review #141

Conversation

ksbhaskar commented Jun 24, 2018

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ksbhaskar left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment