[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [PATCH v5 01/11] qdict: implement a qdict_crumple metho
From: |
Markus Armbruster |
Subject: |
Re: [Qemu-devel] [PATCH v5 01/11] qdict: implement a qdict_crumple method for un-flattening a dict |
Date: |
Thu, 16 Jun 2016 11:16:14 +0200 |
User-agent: |
Gnus/5.13 (Gnus v5.13) Emacs/24.5 (gnu/linux) |
"Daniel P. Berrange" <address@hidden> writes:
> On Thu, Jun 09, 2016 at 03:20:47PM +0200, Markus Armbruster wrote:
>> I apologize for the lateness of this review.
>>
>> "Daniel P. Berrange" <address@hidden> writes:
>>
>> > The qdict_flatten() method will take a dict whose elements are
>> > further nested dicts/lists and flatten them by concatenating
>> > keys.
>> >
>> > The qdict_crumple() method aims to do the reverse, taking a flat
>> > qdict, and turning it into a set of nested dicts/lists. It will
>> > apply nesting based on the key name, with a '.' indicating a
>> > new level in the hierarchy. If the keys in the nested structure
>> > are all numeric, it will create a list, otherwise it will create
>> > a dict.
>> >
>> > If the keys are a mixture of numeric and non-numeric, or the
>> > numeric keys are not in strictly ascending order, an error will
>> > be reported.
>> >
>> > As an example, a flat dict containing
>> >
>> > {
>> > 'foo.0.bar': 'one',
>> > 'foo.0.wizz': '1',
>> > 'foo.1.bar': 'two',
>> > 'foo.1.wizz': '2'
>> > }
>> >
>> > will get turned into a dict with one element 'foo' whose
>> > value is a list. The list elements will each in turn be
>> > dicts.
>> >
>> > {
>> > 'foo': [
>> > { 'bar': 'one', 'wizz': '1' },
>> > { 'bar': 'two', 'wizz': '2' }
>> > ],
>> > }
>> >
>> > If the key is intended to contain a literal '.', then it must
>> > be escaped as '..'. ie a flat dict
>> >
>> > {
>> > 'foo..bar': 'wizz',
>> > 'bar.foo..bar': 'eek',
>> > 'bar.hello': 'world'
>> > }
>> >
>> > Will end up as
>> >
>> > {
>> > 'foo.bar': 'wizz',
>> > 'bar': {
>> > 'foo.bar': 'eek',
>> > 'hello': 'world'
>> > }
>> > }
>> >
>> > The intent of this function is that it allows a set of QemuOpts
>> > to be turned into a nested data structure that mirrors the nesting
>> > used when the same object is defined over QMP.
>> >
>> > Signed-off-by: Daniel P. Berrange <address@hidden>
>>
>
>> > +/**
>> > + * qdict_split_flat_key:
>> > + * @key: the key string to split
>> > + * @prefix: non-NULL pointer to hold extracted prefix
>> > + * @suffix: non-NULL pointer to hold extracted suffix
>> > + *
>> > + * Given a flattened key such as 'foo.0.bar', split it
>> > + * into two parts at the first '.' separator. Allows
>> > + * double dot ('..') to escape the normal separator.
>> > + *
>> > + * eg
>> > + * 'foo.0.bar' -> prefix='foo' and suffix='0.bar'
>> > + * 'foo..0.bar' -> prefix='foo.0' and suffix='bar'
>> > + *
>> > + * The '..' sequence will be unescaped in the returned
>> > + * 'prefix' string. The 'suffix' string will be left
>> > + * in escaped format, so it can be fed back into the
>> > + * qdict_split_flat_key() key as the input later.
>>
>> Why is the suffix strdup'ed then?
>
> It doesn't need to be - i'll make it const.
>
>
>
>> > +}
>> > +
>> > +
>> > +/**
>> > + * qdict_list_size:
>> > + * @maybe_list: dict that may be only list elements
>>
>> Huh? How can a dictionary "be only list elements"?
>>
>> Do you mean "the dictionary to test?"
>
> I'll say "dict to search for keys representing list elements."
Yes, that's better.
>> > + *
>> > + * Determine whether all keys in @maybe_list are
>> > + * valid list elements. If they are all valid,
>> > + * then this returns the number of elements. If
>> > + * they all look like non-numeric keys, then returns
>> > + * zero. If there is a mix of numeric and non-numeric
>> > + * keys, then an error is set as it is both a list
>> > + * and a dict at once.
>>
>> This is well-defined only if empty @maybe_list is considered to have
>> dict nature, not list nature. Else, return value zero could be the
>> length of the empty list or the special "has dict nature" value.
>>
>> Please spell out behavior for empty @maybe_list.
>
> Yep, empty list will imply qdict nature and so return zero.
>
>> > + *
>> > + * Returns: number of list elements, 0 if a dict, -1 on error
>>
>> Awkward function name. qdict_list_size_if_list() would be clear.
>>
>> But I'd simply turn this into a predicate qdict_is_list(), and have the
>> caller use qdict_size() to get the number of elements.
>
> I thought that qdict_size() would cause another iteration, but
> I see now it just returns a cached size. So I'll switch to
> qidct_is_list().
>
>> > +static ssize_t qdict_list_size(QDict *maybe_list, Error **errp)
>> > +{
>> > + const QDictEntry *entry, *next;
>> > + ssize_t len = 0;
>> > + ssize_t max = -1;
>> > + int is_list = -1;
>> > + int64_t val;
>> > +
>> > + entry = qdict_first(maybe_list);
>> > + while (entry != NULL) {
>> > + next = qdict_next(maybe_list, entry);
>>
>> Please keep the loop control in one place:
>>
>> for (entry = qdict_first(maybe_list); entry; entry =
>> qdict_next(entry)) {
>>
>> I'd rename some variables for less verbiage:
>>
>> for (ent = qdict_first(dict); ent; ent = qdict_next(ent)) {
>>
>> Your loop control also works when the loop body deletes @entry from
>> @maybe_list. Seeing such loop control in a function that isn't supposed
>> to change the its argument makes the reviewer go "WTF?!?" :)
>
> This pattern was left from an earlier version where all the code was in
> one method. I'll change it to a for() loop now.
>
>>
>> > +
>> > + if (qemu_strtoll(entry->key, NULL, 10, &val) == 0) {
>> > + if (is_list == -1) {
>> > + is_list = 1;
>> > + } else if (!is_list) {
>> > + error_setg(errp,
>> > + "Cannot crumple a dictionary with a mix of
>> > list "
>> > + "and non-list keys");
>> > + return -1;
>> > + }
>> > + len++;
>> > + if (val > max) {
>> > + max = val;
>> > + }
>> > + } else {
>> > + if (is_list == -1) {
>> > + is_list = 0;
>> > + } else if (is_list) {
>> > + error_setg(errp,
>> > + "Cannot crumple a dictionary with a mix of
>> > list "
>> > + "and non-list keys");
>> > + return -1;
>> > + }
>> > + }
>> > +
>> > + entry = next;
>> > + }
>> > +
>> > + if (is_list == -1) {
>> > + is_list = 0;
>>
>> This can happen only when @maybe_list is empty. Okay, but perhaps you'd
>> like to assert(!qdict_size(maybe_list)).
>
> Ok
>
>>
>> > + }
>> > +
>> > + if (len != (max + 1)) {
>> > + error_setg(errp, "List indexes are not continuous, "
>> > + "saw %zd elements but %zd largest index",
>> > + len, max);
>> > + return -1;
>>
>> contiguous?
>>
>> What if we saw indexes 0, 2, 2?
>
> That would imply that the dict had two entries with the same
> key, which by definition is impossible with a dict data
> structure.
I got confused :)
>> > + }
>> > +
>> > + return is_list ? len : 0;
>> > +}
>> > +
>> > +/**
>> > + * qdict_crumple:
>> > + * @src: the original flat dictionary to crumple
>>
>> "Flat" means all values are scalar. Should we spell that out?
>
> Yep, ok.
>
>> > + * @recursive: true to recursively crumple nested dictionaries
>> > + *
>> > + * Takes a flat dictionary whose keys use '.' separator to
>> > + * indicate nesting, and values are scalars, crumplings it
>>
>> s/, crumplings/, and crumples/
>>
>> > + * into a nested structure. If the @recursive parameter is
>> > + * false, then only the first level of structure implied
>> > + * by the keys will be crumpled. If @recursive is true,
>> > + * then the input will be recursively crumpled to expand
>> > + * all levels of structure in the keys.
>> > + *
>> > + * To include a literal '.' in a key name, it must be escaped
>> > + * as '..'
>> > + *
>> > + * For example, an input of:
>> > + *
>> > + * { 'foo.0.bar': 'one', 'foo.0.wizz': '1',
>> > + * 'foo.1.bar': 'two', 'foo.1.wizz': '2' }
>> > + *
>> > + * will result in any output of:
>> > + *
>> > + * {
>> > + * 'foo': [
>> > + * { 'bar': 'one', 'wizz': '1' },
>> > + * { 'bar': 'two', 'wizz': '2' }
>> > + * ],
>> > + * }
>> > + *
>> > + * Returns: either a QDict or QList for the nested data structure
>>
>> I think you should discuss how this can fail.
>
> Will do.
>
>> > +QObject *qdict_crumple(QDict *src, bool recursive, Error **errp)
>> > +{
>> > + const QDictEntry *entry, *next;
>> > + QDict *two_level, *multi_level = NULL;
>> > + QObject *dst = NULL, *child;
>> > + ssize_t list_len;
>> > + size_t i;
>> > + char *prefix = NULL, *suffix = NULL;
>> > +
>> > + two_level = qdict_new();
>> > + entry = qdict_first(src);
>> > +
>> > + /* Step 1: split our totally flat dict into a two level dict */
>> > + while (entry != NULL) {
>> > + next = qdict_next(src, entry);
>>
>> Again, keep the loop control in one place.
>>
>> > +
>> > + if (qobject_type(entry->value) == QTYPE_QDICT ||
>> > + qobject_type(entry->value) == QTYPE_QLIST) {
>> > + error_setg(errp, "Value %s is not a scalar",
>> > + entry->key);
>> > + goto error;
>> > + }
>> > +
>> > + qdict_split_flat_key(entry->key, &prefix, &suffix);
>> > +
>> > + child = qdict_get(two_level, prefix);
>> > + if (suffix) {
>> > + if (child) {
>> > + if (qobject_type(child) != QTYPE_QDICT) {
>> > + error_setg(errp, "Key %s prefix is already set as a
>> > scalar",
>> > + prefix);
>> > + goto error;
>> > + }
>> > + } else {
>> > + child = QOBJECT(qdict_new());
>> > + qdict_put_obj(two_level, prefix, child);
>> > + }
>> > + qobject_incref(entry->value);
>> > + qdict_put_obj(qobject_to_qdict(child), suffix, entry->value);
>> > + } else {
>> > + if (child) {
>> > + error_setg(errp, "Key %s prefix is already set as a dict",
>> > + prefix);
>> > + goto error;
>> > + }
>> > + qobject_incref(entry->value);
>> > + qdict_put_obj(two_level, prefix, entry->value);
>> > + }
>>
>> Works, because we put only QDicts we've created ourselves (first
>> qdict_put_obj() above) and values we got from @src (second
>> qdict_put_obj()), and we fail when such a value isn't scalar.
>>
>> > +
>> > + g_free(suffix);
>>
>> As I suspected, qdict_split_flat_key() strdup'ing the suffix is useless.
>>
>> > + g_free(prefix);
>> > + suffix = prefix = NULL;
>>
>> Dead stores.
>
> No, they're not dead. The end of this method has a 'g_free(prefix)' so
> we must be sure to clear this out to prevent a double-free if a later
> codebase jumps to the error label.
>
>> > + entry = next;
>> > + }
>> > +
>> > + /* Step 2: optionally process the two level dict recursively
>> > + * into a multi-level dict */
>> > + if (recursive) {
>> > + multi_level = qdict_new();
>> > + entry = qdict_first(two_level);
>> > + while (entry != NULL) {
>> > + next = qdict_next(two_level, entry);
>>
>> Again, keep the loop control in one place.
>
> OK
>
>>
>> > +
>> > + if (qobject_type(entry->value) == QTYPE_QDICT) {
>> > + child = qdict_crumple(qobject_to_qdict(entry->value),
>> > + recursive, errp);
>> > + if (!child) {
>> > + goto error;
>> > + }
>> > +
>> > + qdict_put_obj(multi_level, entry->key, child);
>> > + } else {
>> > + qobject_incref(entry->value);
>> > + qdict_put_obj(multi_level, entry->key, entry->value);
>> > + }
>> > +
>> > + entry = next;
>> > + }
>> > + QDECREF(two_level);
>> > + } else {
>> > + multi_level = two_level;
>> > + }
>> > + two_level = NULL;
>> > +
>> > + /* Step 3: detect if we need to turn our dict into list */
>> > + list_len = qdict_list_size(multi_level, errp);
>> > + if (list_len < 0) {
>> > + goto error;
>> > + }
>> > +
>> > + if (list_len) {
>> > + dst = QOBJECT(qlist_new());
>> > +
>> > + for (i = 0; i < list_len; i++) {
>> > + char *key = g_strdup_printf("%zu", i);
>> > +
>> > + child = qdict_get(multi_level, key);
>> > + g_free(key);
>> > + assert(child);
>>
>> qdict_list_size() accepts as list index any (string) key qemu_strtoll()
>> accepts. If %zu formats it back into the same string, we'll find it
>> here. Else we die. Please spell this out in the function contract.
>
> OK.
>
>> [*] I'm afraid we also die if qdict_list_size()'s "List indexes are not
>> continuous" check gets fooled. Suggest to drop that check, and replace
>> this assertion by error_setg(errp, "Malformed list indexes").
>> Admittedly not the nicest error message; perhaps you can come up with a
>> better one.
>
> We can't be fooled per my note earlier
>
>>
>> > +
>> > + qobject_incref(child);
>> > + qlist_append_obj(qobject_to_qlist(dst), child);
>> > + }
>> > + QDECREF(multi_level);
>>
>> Do we need
>>
>> multi_level = NULL;
>>
>> here?
>
> It isn't needed right now, since we're at the end of the method now
> and nothing will touch this var again. Setting it to NULL is a
> worthwhile safety net for future refactoring though.
I generally don't bother to zap pointers "just in case". I saw the
QDECREF() after the error label, and missed the fact that we can't reach
it from here. Your patch is fine as is.
>> > + } else {
>> > + dst = QOBJECT(multi_level);
>> > + }
>> > +
>> > + return dst;
>> > +
>> > + error:
>> > + g_free(suffix);
>> > + g_free(prefix);
>> > + QDECREF(multi_level);
>> > + QDECREF(two_level);
>> > + qobject_decref(dst);
>> > + return NULL;
>> > +}
>> > +
>> > +
>> > /**
>> > * qdict_array_entries(): Returns the number of direct array entries if
>> > the
>> > * sub-QDict of src specified by the prefix in subqdict (or src itself for
>> > diff --git a/tests/check-qdict.c b/tests/check-qdict.c
>> > index a43056c..0d12f40 100644
>> > --- a/tests/check-qdict.c
>> > +++ b/tests/check-qdict.c
>> > @@ -15,6 +15,7 @@
>> > #include "qapi/qmp/qint.h"
>> > #include "qapi/qmp/qdict.h"
>> > #include "qapi/qmp/qstring.h"
>> > +#include "qapi/error.h"
>> > #include "qemu-common.h"
>
>> > +static void qdict_crumple_test_bad_inputs(void)
>> > +{
>> > + QDict *src;
>> > + Error *error = NULL;
>> > +
>> > + src = qdict_new();
>> > + /* rule.0 can't be both a string and a dict */
>> > + qdict_put(src, "rule.0", qstring_from_str("fred"));
>> > + qdict_put(src, "rule.0.policy", qstring_from_str("allow"));
>> > +
>> > + g_assert(qdict_crumple(src, true, &error) == NULL);
>> > + g_assert(error != NULL);
>> > + error_free(error);
>> > + error = NULL;
>> > + QDECREF(src);
>> > +
>> > + src = qdict_new();
>> > + /* rule can't be both a list and a dict */
>> > + qdict_put(src, "rule.0", qstring_from_str("fred"));
>> > + qdict_put(src, "rule.a", qstring_from_str("allow"));
>> > +
>> > + g_assert(qdict_crumple(src, true, &error) == NULL);
>> > + g_assert(error != NULL);
>> > + error_free(error);
>> > + error = NULL;
>> > + QDECREF(src);
>> > +
>> > + src = qdict_new();
>> > + /* The input should be flat, ie no dicts or lists */
>> > + qdict_put(src, "rule.a", qdict_new());
>> > + qdict_put(src, "rule.b", qstring_from_str("allow"));
>> > +
>> > + g_assert(qdict_crumple(src, true, &error) == NULL);
>> > + g_assert(error != NULL);
>> > + error_free(error);
>> > + error = NULL;
>> > + QDECREF(src);
>> > +
>> > +
>> > + src = qdict_new();
>> > + /* List indexes must not have gaps */
>> > + qdict_put(src, "rule.0", qdict_new());
>> > + qdict_put(src, "rule.3", qstring_from_str("allow"));
>> > +
>> > + g_assert(qdict_crumple(src, true, &error) == NULL);
>> > + g_assert(error != NULL);
>> > + error_free(error);
>> > + error = NULL;
>> > + QDECREF(src);
>>
>> Suggest to add test case
>>
>> /* List indexes must not have gaps (more creative version) */
>> qdict_put(src, "rule.0", ...);
>> qdict_put(src, "rule.2", ...);
>> qdict_put(src, "rule.2", ...);
>
> That's surely impossible, as dict keys have to be unique.
It's surely possible that patch review melts my brain :)
>> and
>>
>> /* List indexes must be in %zu format */
>> qdict_put(src, "rule.+0", ...);
- [Qemu-devel] [PATCH v5 00/11] Provide a QOM-based authorization API, Daniel P. Berrange, 2016/06/02
- [Qemu-devel] [PATCH v5 04/11] util: add QAuthZ object as an authorization base class, Daniel P. Berrange, 2016/06/02
- [Qemu-devel] [PATCH v5 03/11] qom: support arbitrary non-scalar properties with -object, Daniel P. Berrange, 2016/06/02
- [Qemu-devel] [PATCH v5 05/11] util: add QAuthZSimple object type for a simple access control list, Daniel P. Berrange, 2016/06/02
- [Qemu-devel] [PATCH v5 02/11] qapi: allow QmpInputVisitor to auto-cast types, Daniel P. Berrange, 2016/06/02