-#include "grob.hh"
+#include <cstdio>
+#include <cstdlib>
+
#include "item.hh"
-#include "spanner.hh"
-#include "system.hh"
+#include "system.hh"
-static SCM break_criterion;
+static SCM break_criterion;
void
set_break_subsititution (SCM criterion)
{
}
/*
- Perform the substitution for a single grob.
- */
+ Perform the substitution for a single grob.
+*/
SCM
substitute_grob (Grob *sc)
{
- if (SCM_INUMP (break_criterion))
+ if (scm_is_integer (break_criterion))
{
- Item * i = dynamic_cast<Item*> (sc);
+ Item *i = dynamic_cast<Item *> (sc);
Direction d = to_dir (break_criterion);
if (i && i->break_status_dir () != d)
{
}
else
{
- System * line
- = dynamic_cast<System*> (unsmob_grob (break_criterion));
+ System *line
+ = dynamic_cast<System *> (unsmob_grob (break_criterion));
if (sc->get_system () != line)
{
sc = sc->find_broken_piece (line);
-
}
-
+
/* now: !sc || (sc && sc->get_system () == line) */
if (!sc)
return SCM_UNDEFINED;
/* now: sc && sc->get_system () == line */
if (!line)
- return sc->self_scm();
+ return sc->self_scm ();
/*
We don't return SCM_UNDEFINED for
- it forces us to mark the parents of a grob, leading to
a huge recursion in the GC routine.
- */
-
- /*
- This was introduced in 1.3.49 as a measure to prevent
- programming errors. It looks rather expensive (?).
-
- TODO:
-
- benchmark , document when (what kind of programming
- errors) this happens.
*/
+
if (sc->common_refpoint (line, X_AXIS)
&& sc->common_refpoint (line, Y_AXIS))
{
return SCM_UNDEFINED;
}
- return sc->self_scm();
+ return sc->self_scm ();
}
-
-
/*
Do break substitution in S, using CRITERION. Return new value.
CRITERION is either a SMOB pointer to the desired line, or a number
do_break_substitution (SCM src)
{
again:
-
+
if (unsmob_grob (src))
+ return substitute_grob (unsmob_grob (src));
+ else if (scm_is_vector (src))
{
- return substitute_grob (unsmob_grob (src));
- }
- else if (gh_vector_p (src))
- {
- int l = SCM_VECTOR_LENGTH (src);
- SCM nv = scm_c_make_vector (l, SCM_UNDEFINED);
-
- for (int i =0 ; i< l ; i++)
+ int len = scm_c_vector_length (src);
+ SCM nv = scm_c_make_vector (len, SCM_UNDEFINED);
+ for (int i = 0; i < len; i++)
{
- SCM si = gh_int2scm (i);
- scm_vector_set_x (nv, si, do_break_substitution (scm_vector_ref (src, si)));
+ SCM si = scm_int2num (i);
+ scm_vector_set_x (nv, si,
+ do_break_substitution (scm_vector_ref (src, si)));
}
}
- else if (ly_pair_p (src))
+ else if (scm_is_pair (src))
{
/*
UGH! breaks on circular lists.
*/
- SCM newcar = do_break_substitution (ly_car (src));
- SCM oldcdr = ly_cdr (src);
-
+ SCM newcar = do_break_substitution (scm_car (src));
+ SCM oldcdr = scm_cdr (src);
+
if (newcar == SCM_UNDEFINED
- && (gh_pair_p (oldcdr) || oldcdr == SCM_EOL))
+ && (scm_is_pair (oldcdr) || oldcdr == SCM_EOL))
{
/*
- This is tail-recursion, ie.
-
+ This is tail-recursion, ie.
+
return do_break_substution (cdr);
We don't want to rely on the compiler to do this. Without
tail-recursion, this easily crashes with a stack overflow. */
- src = oldcdr;
+ src = oldcdr;
goto again;
}
return src;
}
-
/*
Perform substitution on GROB_LIST using a constant amount of stack.
- */
+*/
SCM
substitute_grob_list (SCM grob_list)
{
SCM l = SCM_EOL;
- SCM * tail = &l;
+ SCM *tail = &l;
- for (SCM s = grob_list; gh_pair_p (s); s = gh_cdr (s))
+ for (SCM s = grob_list; scm_is_pair (s); s = scm_cdr (s))
{
- SCM n= substitute_grob (unsmob_grob (gh_car (s)));
+ SCM n = substitute_grob (unsmob_grob (scm_car (s)));
if (n != SCM_UNDEFINED)
{
- *tail = gh_cons (n, SCM_EOL);
- tail = SCM_CDRLOC(*tail);
+ *tail = scm_cons (n, SCM_EOL);
+ tail = SCM_CDRLOC (*tail);
}
}
return l;
}
+/*
+ We don't do
+
+ forall b in broken-childs:
+ forall p in properties:
+ forall g in p (if grob-list):
+ g := substitute (g)
+
+ for spanners since this is O (SYSTEMCOUNT * GROBCOUNT), and SYSTEMCOUNT =
+ O (GROBCOUNT), we have a quadratic algorithm. --for a single spanner
+
+ This is problematic: with large (long) scores, the costs can be
+ significant; especially all-elements in System, can become huge. For
+ a typical 50 page score, it requires running through a 100k list 50
+ times.
+
+ Instead:
+
+ forall p in properties:
+ (if grob list)
+
+ put grob list in array,
+
+ reorder array so spanners are separate -- O (grobcount)
+
+ find first and last indexes of grobs on a specific system
+
+ for items this is O (itemcount)
+
+ for spanners this is O (sum-of spanner-system-ranges)
+
+ perform the substitution O (sum-of spanner-system-ranges)
+
+
+ The complexity is harder to determine, but should be subquadratic;
+
+ For the situation above, we run through the entire 100k list once,
+ and also (more or less) once through the item part of the 100k (say
+ 98k elements) of the list.
+
+
+ These timings were measured without -O2.
+
+ lehre, before 28.98 seconds, after: 27.91 seconds, 3.5 %.
+
+ coriolan, before 2:30, after: 1:59. Increase of 20%.
+
+ moz-k498-p1, before 24.10, after: 19.790s, Increase of 18%
+*/
+
+Slice
+spanner_system_range (Spanner *sp)
+{
+ Slice rv;
+
+ if (System *st = sp->get_system ())
+ {
+ rv = Slice (st->rank_, st->rank_);
+ }
+ else
+ {
+ if (sp->broken_intos_.size ())
+ rv = Slice (sp->broken_intos_[0]->get_system ()->rank_,
+ sp->broken_intos_.top ()->get_system ()->rank_);
+ }
+ return rv;
+}
+
+Slice
+item_system_range (Item *it)
+{
+ if (System *st = it->get_system ())
+ return Slice (st->rank_, st->rank_);
+
+ Slice sr;
+ Direction d = LEFT;
+ do
+ {
+ Item *bi = it->find_prebroken_piece (d);
+ if (bi && bi->get_system ())
+ sr.add_point (bi->get_system ()->rank_);
+ }
+ while (flip (&d) != LEFT);
+
+ return sr;
+}
+
+Slice
+grob_system_range (Grob *g)
+{
+ if (Spanner *s = dynamic_cast<Spanner *> (g))
+ return spanner_system_range (s);
+ else if (Item *it = dynamic_cast<Item *> (g))
+ return item_system_range (it);
+ else
+ return Slice ();
+}
+
+struct Substitution_entry
+{
+ Grob *grob_;
+ short left_;
+ short right_;
+
+ void set (Grob *g, Slice sr)
+ {
+ grob_ = g;
+ /*
+ duh, don't support scores with more than 32000 systems.
+ */
+ if (sr.is_empty ())
+ {
+ /*
+ overflow if we don't treat this specially.
+ */
+ left_ = 1;
+ right_ = -1;
+ }
+ else
+ {
+ left_ = sr[LEFT];
+ right_ = sr[RIGHT];
+ }
+ }
+ Substitution_entry ()
+ {
+ grob_ = 0;
+ left_ = right_ = -2;
+ }
+
+ int length () { return right_ - left_; }
+ static int
+ item_compare (void const *a, void const *b)
+ {
+ return ((Substitution_entry *)a)->left_
+ - ((Substitution_entry *)b)->left_;
+ }
+
+ static int
+ spanner_compare (void const *a, void const *b)
+ {
+ return ((Substitution_entry *)a)->length ()
+ - ((Substitution_entry *)b)->length ();
+ }
+};
+
+bool
+Spanner::fast_fubstitute_grob_list (SCM sym,
+ SCM grob_list)
+{
+ int len = scm_ilength (grob_list);
+
+ /*
+ Only do this complicated thing for large lists. This has the added
+ advantage that we won't screw up the ordering for elements in
+ alignments (which typically don't have more than 10 grobs.)
+ */
+
+ if (len < 300)
+ return false;
-SCM grob_list_p;
+ /*
+ TODO : should not free it some time?
+ */
+ static Substitution_entry *vec;
+ static int vec_room;
+
+ if (vec_room < len)
+ {
+ vec = (Substitution_entry *) realloc (vec, sizeof (Substitution_entry) * len);
+ vec_room = len;
+ }
+
+ Slice system_range = spanner_system_range (this);
+
+ Array<Slice> it_indices;
+ Array<Slice> sp_indices;
+ for (int i = 0; i <= system_range.length (); i++)
+ {
+ it_indices.push (Slice (len, 0));
+ sp_indices.push (Slice (len, 0));
+ }
+
+ int sp_index = len;
+ int it_index = 0;
+ for (SCM s = grob_list; scm_is_pair (s); s = scm_cdr (s))
+ {
+ Grob *g = unsmob_grob (scm_car (s));
+
+ Slice sr = grob_system_range (g);
+ sr.intersect (system_range);
+
+ int idx = 0;
+ if (dynamic_cast<Spanner *> (g))
+ {
+ idx =--sp_index;
+ }
+ else if (dynamic_cast<Item *> (g))
+ {
+ idx = it_index++;
+ }
+
+ vec[idx].set (g, sr);
+ }
+
+ qsort (vec, it_index,
+ sizeof (Substitution_entry), &Substitution_entry::item_compare);
+
+ Array<Slice> *arrs[]
+ = {
+ &it_indices, &sp_indices
+ };
+
+ for (int i = 0; i < it_index;i++)
+ {
+ for (int j = vec[i].left_; j <= vec[i].right_; j++)
+ {
+ it_indices[j - system_range[LEFT]].add_point (i);
+ }
+ }
+
+ /*
+ sorting vec[sp_index.. len]
+ is a waste of time -- the staff-spanners screw up the
+ ordering, since they go across the entire score.
+ */
+ for (int i = sp_indices.size (); i--;)
+ sp_indices[i] = Slice (sp_index, len - 1);
+
+ assert (it_index <= sp_index);
+
+ assert (broken_intos_.size () == system_range.length () + 1);
+ for (int i = 0; i < broken_intos_.size (); i++)
+ {
+ Grob *sc = broken_intos_[i];
+ System *l = sc->get_system ();
+ set_break_subsititution (l ? l->self_scm () : SCM_UNDEFINED);
+
+ SCM newval = SCM_EOL;
+ SCM *tail = &newval;
+
+ for (int k = 0; k < 2;k++)
+ for (int j = (*arrs[k])[i][LEFT]; j <= (*arrs[k])[i][RIGHT]; j++)
+ {
+ SCM subs = substitute_grob (vec[j].grob_);
+ if (subs != SCM_UNDEFINED)
+ {
+ *tail = scm_cons (subs, SCM_EOL);
+
+ tail = SCM_CDRLOC (*tail);
+ }
+ }
+
+#ifdef PARANOIA
+
+ printf ("%d (%d), sp %d (%d)\n",
+ it_indices [i].length (), it_index,
+ sp_indices[i].length (), len -sp_index);
+
+ {
+ SCM l1 = substitute_grob_list (grob_list);
+ assert (scm_ilength (l1) == scm_ilength (newval));
+ }
+#endif
+
+ /*
+ see below.
+ */
+ if (sym == ly_symbol2scm ("all-elements"))
+ sc->mutable_property_alist_
+ = scm_assq_remove_x (sc->mutable_property_alist_,
+ ly_symbol2scm ("all-elements"));
+
+ sc->mutable_property_alist_ = scm_acons (sym, newval,
+ sc->mutable_property_alist_);
+ }
+
+ return true;
+}
/*
Although the substitution can be written as
we have a special function here: we want to invoke a special
function for lists of grobs. These can be very long for large
- orchestral scores (eg. 1M elements). do_break_substitution() can
+ orchestral scores (eg. 1M elements). do_break_substitution () can
recurse many levels, taking lots of stack space.
This becomes a problem if lily is linked against guile with
pthreads. pthreads impose small limits on the stack size.
- */
+*/
SCM
substitute_mutable_property_alist (SCM alist)
{
- if (!grob_list_p)
- grob_list_p = scm_c_eval_string ("grob-list?");
+ SCM grob_list_p = ly_lily_module_constant ("grob-list?");
SCM l = SCM_EOL;
SCM *tail = &l;
- for (SCM s = alist; gh_pair_p (s); s = gh_cdr (s))
+ for (SCM s = alist; scm_is_pair (s); s = scm_cdr (s))
{
- SCM sym = gh_caar(s);
- SCM val = gh_cdar(s);
+ SCM sym = scm_caar (s);
+ SCM val = scm_cdar (s);
SCM type = scm_object_property (sym, ly_symbol2scm ("backend-type?"));
if (type == grob_list_p)
else
val = do_break_substitution (val);
- *tail = gh_cons (gh_cons (sym, val), SCM_EOL);
- tail = SCM_CDRLOC (*tail);
+ if (val != SCM_UNDEFINED)
+ {
+ /*
+ for ly:grob? properties, SCM_UNDEFINED could leak out
+ through ly:grob-property
+ */
+ *tail = scm_cons (scm_cons (sym, val), SCM_EOL);
+ tail = SCM_CDRLOC (*tail);
+ }
}
-
return l;
}
+void
+Spanner::substitute_one_mutable_property (SCM sym,
+ SCM val)
+{
+ SCM type = scm_object_property (sym, ly_symbol2scm ("backend-type?"));
+ Spanner *s = this;
+
+ bool fast_done = false;
+ SCM grob_list_p = ly_lily_module_constant ("grob-list?");
+ if (type == grob_list_p)
+ fast_done = s->fast_fubstitute_grob_list (sym, val);
+
+ if (!fast_done)
+ for (int i = 0; i < s->broken_intos_.size (); i++)
+ {
+ Grob *sc = s->broken_intos_[i];
+ System *l = sc->get_system ();
+ set_break_subsititution (l ? l->self_scm () : SCM_UNDEFINED);
+
+ SCM newval = (type == grob_list_p)
+ ? substitute_grob_list (val)
+ : do_break_substitution (val);
+
+ /*
+ For the substitution of a single property, we tack the result onto
+ mutable_property_alist_ ; mutable_property_alist_ is empty after
+ Grob::Grob (Grob const&), except that System has all-elements set,
+ as a side product of typeset_grob () on newly copied spanners.
+
+ Here we clear that list explicitly to free some memory and
+ counter some of the confusion I encountered while debugging
+ another problem
+
+ (hwn 4/2/04)
+ */
+ if (sym == ly_symbol2scm ("all-elements"))
+ sc->mutable_property_alist_
+ = scm_assq_remove_x (sc->mutable_property_alist_,
+ ly_symbol2scm ("all-elements"));
+
+ sc->mutable_property_alist_ = scm_cons (scm_cons (sym, newval),
+ sc->mutable_property_alist_);
+ }
+}