Scans for {{{token}}} (triple-brace, unescaped) occurrences: onepagr
templates never use double-brace {{token}} (double-brace HTML-escapes
and corrupts any value containing "&", "<", or ">"). Section markers
and comments are not matched: onepagr templates use flat
triple-brace substitution only, no Mustache sections or partials.
Details
// line comments are stripped before scanning: Typst templates
routinely document the triple-brace convention with a literal
{{{token}}} example in a header comment (this is a real case, not
hypothetical; the reference trend-snapshot template does exactly
this), and without stripping comments first, that illustrative example
is indistinguishable from a real required token. // is unambiguously
a comment marker in Typst (division is a single /), so this is safe
for any Typst source: the one caveat is a literal // inside a
string constant in the template's own code (not data, which arrives
via tokens), which onepagr's built-in templates never do.
A token with a declared default (see the optional-token marker in
compile_typst()) is not required, so it is left out of the result.
A default can refer to other tokens ({{{token}}}); those are required.
Examples
path <- resolve_template("cohort_summary")
extract_required_tokens(path)
#> [1] "contact_email" "logo_primary_path"
#> [3] "logo_primary_alt" "org_full"
#> [5] "contact_url" "doc_title"
#> [7] "doc_subtitle" "strip_data"
#> [9] "strip_period" "strip_design"
#> [11] "strip_geography" "n_decedents"
#> [13] "n_ems_total" "pct_linked"
#> [15] "n_prior_ems" "n_prior_od_ems"
#> [17] "pct_linked_male_width" "n_male"
#> [19] "pct_linked_female_width" "n_female"
#> [21] "pct_linked_appalachian_width" "n_appalachian"
#> [23] "pct_linked_nonappalachian_width" "n_nonappalachian"
#> [25] "pct_linked_white_width" "n_white"
#> [27] "pct_linked_black_width" "n_black"
#> [29] "pct_linked_other_width" "n_other"
#> [31] "pct_linked_age_lt25_width" "n_age_lt25"
#> [33] "pct_linked_age_25_34_width" "n_age_25_34"
#> [35] "pct_linked_age_35_44_width" "n_age_35_44"
#> [37] "pct_linked_age_45_54_width" "n_age_45_54"
#> [39] "pct_linked_age_55_64_width" "n_age_55_64"
#> [41] "pct_linked_age_65plus_width" "n_age_65plus"
#> [43] "tl1_yr" "tl1_label"
#> [45] "tl2_yr" "tl2_label"
#> [47] "tl3_yr" "tl3_label"
#> [49] "pct_any_prior_enc" "pct_od_prior_enc"
#> [51] "pct_naloxone_width" "pct_no_naloxone_width"
#> [53] "pct_naloxone" "n_naloxone_enc"
#> [55] "pct_no_naloxone" "n_no_naloxone_enc"
#> [57] "domain_diff_scene" "domain_diff_history"
#> [59] "domain_diff_drug" "domain_diff_medication"
#> [61] "domain_diff_mental" "median_days"
#> [63] "mean_days" "pct_30d_width"
#> [65] "n_30d" "pct_90d_width"
#> [67] "n_90d" "pct_365d_width"
#> [69] "n_365d" "pct_gt365d_width"
#> [71] "n_gt365d" "lessons_learned_text"
#> [73] "disclaimer_text" "footnote_sources"
#> [75] "n_od_ems_denom" "timing_denom"
#> [77] "n_eligible_decedents" "n_unlinked_decedents"
#> [79] "wc_linked_median" "wc_unlinked_median"
#> [81] "mean_prior_enc" "median_prior_enc"
#> [83] "pct_decedent_nax" "timing_iqr"